Array CGH is a recently introduced technology that measures changes in the gene copy number of hundreds of genes in a single experiment. The primary goal of this study was to develop machine learning models that classify non-small Lung Cancers according to histopathology types and to compare several machine learning methods in this learning task. DNA from tumors of 37 patients (21 squamous carcinomas, and 16 adenocarcinomas) were extracted and hybridized onto a 452 BAC clone array. The following algorithms were used: KNN, Decision Tree Induction, Support Vector Machines and Feed-Forward Neural Networks. Performance was measured via leave-one-out classification accuracy. The best multi-gene model found had a leave-one-out accuracy of 89.2%. Decision Trees performed poorer than the other methods in this learning task and dataset. We conclude that gene copy numbers as measured by array CGH are, collectively, an excellent indicator of histological subtype. Several interesting research directions are discussed.
{"title":"Machine learning models for lung cancer classification using array comparative genomic hybridization.","authors":"C F Aliferis, D Hardin, P P Massion","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>Array CGH is a recently introduced technology that measures changes in the gene copy number of hundreds of genes in a single experiment. The primary goal of this study was to develop machine learning models that classify non-small Lung Cancers according to histopathology types and to compare several machine learning methods in this learning task. DNA from tumors of 37 patients (21 squamous carcinomas, and 16 adenocarcinomas) were extracted and hybridized onto a 452 BAC clone array. The following algorithms were used: KNN, Decision Tree Induction, Support Vector Machines and Feed-Forward Neural Networks. Performance was measured via leave-one-out classification accuracy. The best multi-gene model found had a leave-one-out accuracy of 89.2%. Decision Trees performed poorer than the other methods in this learning task and dataset. We conclude that gene copy numbers as measured by array CGH are, collectively, an excellent indicator of histological subtype. Several interesting research directions are discussed.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244172/pdf/procamiasymp00001-0048.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22139762","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In intensive care physiological variables of the critically ill are measured and recorded in short time intervals. The proper extraction and interpretation of the information contained in this flood of information can hardly be done by experience alone. Intelligent alarm systems are needed to provide suitable bedside decision support. So far there is no commonly accepted standard for detecting the actual clinical state from the patient record. We use the statistical methodology of graphical models based on partial correlations for detecting time-varying relationships between physiological variables. Graphical models provide information on the relationships among physiological variables that is helpful e.g. for variable selection. Separate analyses for different pathophysiological states show that distinct clinical states are characterized by distinct partial correlation structures. Hence, this technique can provide new insights into physiological mechanisms.
{"title":"Detecting relationships between physiological variables using graphical models.","authors":"Michael Imhoff, Ronald Fried, Ursula Gather","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>In intensive care physiological variables of the critically ill are measured and recorded in short time intervals. The proper extraction and interpretation of the information contained in this flood of information can hardly be done by experience alone. Intelligent alarm systems are needed to provide suitable bedside decision support. So far there is no commonly accepted standard for detecting the actual clinical state from the patient record. We use the statistical methodology of graphical models based on partial correlations for detecting time-varying relationships between physiological variables. Graphical models provide information on the relationships among physiological variables that is helpful e.g. for variable selection. Separate analyses for different pathophysiological states show that distinct clinical states are characterized by distinct partial correlation structures. Hence, this technique can provide new insights into physiological mechanisms.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244539/pdf/procamiasymp00001-0381.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22139773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nestor J Rodriguez, Viviam Murillo, José A Borges, Johanna Ortiz, Daniel Z Sands
The user interface of an electronic patient record system can significantly improve user acceptance and ease its adoption process. The design of a user interface should take into consideration the characteristics and the needs of the user incorporating usability engineering principles in the lifecycle of its development. In this paper we describe a study of physician interaction with a paper-based patient record system and a graphical-based electronic patient record system. The usability attributes of learnability, efficiency and satisfaction are evaluated on the whole spectrum of physicians' activities with patient record systems. The results of the study did not reveal a significant difference in the overall time to complete typical physician tasks. However, on average physicians can perform viewing tasks faster, documenting tasks slower and ordering tasks at about the same speed on the graphical-based system than on the paper based system. Physicians were found to be significantly more satisfied with the graphical-based system than with the paper-based system. The results also revealed that physicians with higher levels of computer literacy and typing skills can complete typical tasks in significantly less time on a graphical-based system than physicians with lower levels of computer literacy and typing skills.
{"title":"A usability study of physicians interaction with a paper-based patient record system and a graphical-based electronic patient record system.","authors":"Nestor J Rodriguez, Viviam Murillo, José A Borges, Johanna Ortiz, Daniel Z Sands","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>The user interface of an electronic patient record system can significantly improve user acceptance and ease its adoption process. The design of a user interface should take into consideration the characteristics and the needs of the user incorporating usability engineering principles in the lifecycle of its development. In this paper we describe a study of physician interaction with a paper-based patient record system and a graphical-based electronic patient record system. The usability attributes of learnability, efficiency and satisfaction are evaluated on the whole spectrum of physicians' activities with patient record systems. The results of the study did not reveal a significant difference in the overall time to complete typical physician tasks. However, on average physicians can perform viewing tasks faster, documenting tasks slower and ordering tasks at about the same speed on the graphical-based system than on the paper based system. Physicians were found to be significantly more satisfied with the graphical-based system than with the paper-based system. The results also revealed that physicians with higher levels of computer literacy and typing skills can complete typical tasks in significantly less time on a graphical-based system than physicians with lower levels of computer literacy and typing skills.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244463/pdf/procamiasymp00001-0708.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22139780","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Weaning from mechanical ventilation is the gradual detachment from any ventilatory support till normal spontaneous breathing can be fully resumed. To date, we have developed a fuzzy logic controller for weaning COPD adults using pressure support ventilation (PS). However, adults and newborns differ in the pathophysiology of lung disease. We therefore used our fuzzy logic-based weaning platform to develop modularized components for weaning newborns with lung disease. Our controller uses the heart rate (HR), respiratory rate (RR), tidal volume (VT) and oxygen saturation (SaO2) and their trends deltaHR/deltat, deltaVT/deltat and deltaSaO2/deltat to evaluate, respectively, the Current and Trend weaning status of the newborn. Through appropriate fuzzification of these vital signs, Current and Trend weaning status can quantitatively determine the increase/decrease in the synchronized intermittent mandatory ventilation (SIMV) setting. The post-operative weaning courses of 10 newborns, 82+/-162 days old, were assessed at 2-hour intervals for 68+/-39 days. The SIMV levels, proposed by our algorithm, were matched to those levels actually applied. For 60% of the time both values coincided. For the remaining 40%, our algorithm suggested lower SIMV support than what was applied. The Area Under the Curve for integrated ventilatory support over time was 1203+/-846 for standard ventilatory strategies and 1152+/-802 for fuzzy controller. This suggests that the algorithm, approximates the actual weaning progression, and may advocate a more aggressive strategy. Moreover, the core of the fuzzy controller facilitates adaptation for body size and diversified disease patterns and sets the premises as an infant-weaning tool.
{"title":"Fuzzy logic controller for weaning neonates from mechanical ventilation.","authors":"G E Hatzakis, G M Davis","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>Weaning from mechanical ventilation is the gradual detachment from any ventilatory support till normal spontaneous breathing can be fully resumed. To date, we have developed a fuzzy logic controller for weaning COPD adults using pressure support ventilation (PS). However, adults and newborns differ in the pathophysiology of lung disease. We therefore used our fuzzy logic-based weaning platform to develop modularized components for weaning newborns with lung disease. Our controller uses the heart rate (HR), respiratory rate (RR), tidal volume (VT) and oxygen saturation (SaO2) and their trends deltaHR/deltat, deltaVT/deltat and deltaSaO2/deltat to evaluate, respectively, the Current and Trend weaning status of the newborn. Through appropriate fuzzification of these vital signs, Current and Trend weaning status can quantitatively determine the increase/decrease in the synchronized intermittent mandatory ventilation (SIMV) setting. The post-operative weaning courses of 10 newborns, 82+/-162 days old, were assessed at 2-hour intervals for 68+/-39 days. The SIMV levels, proposed by our algorithm, were matched to those levels actually applied. For 60% of the time both values coincided. For the remaining 40%, our algorithm suggested lower SIMV support than what was applied. The Area Under the Curve for integrated ventilatory support over time was 1203+/-846 for standard ventilatory strategies and 1152+/-802 for fuzzy controller. This suggests that the algorithm, approximates the actual weaning progression, and may advocate a more aggressive strategy. Moreover, the core of the fuzzy controller facilitates adaptation for body size and diversified disease patterns and sets the premises as an infant-weaning tool.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244473/pdf/procamiasymp00001-0356.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22139848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rich Caruana, Radu S Niculescu, R Bharat Rao, Cynthia Simms
We apply machine learning to the problem of subpopulation assessment for Caesarian Section. In subpopulation assessment, we are interested in making predictions not for a single patient, but for groups of patients. Typically, in any large population, different subpopulations will have different "outcome" rates. In our example, the C-section rate of a population of 22,176 expectant mothers is 16.8%; yet, the 17 physician groups that serve this population have vastly different group C-section rates, ranging from 11% to 23%. The ultimate goal of subpopulation assessment is to determine if these variations in the observed rates can be attributed to (a) variations in intrinsic risk of the patient sub-populations (i.e. some groups contain more "high-risk C-section" patients), or (b) differences in physician practice (i.e. some groups do more C-sections). Our results indicate that although there is some variation in intrinsic risk, there is also much variation in physician practice.
{"title":"Machine learning for sub-population assessment: evaluating the C-section rate of different physician practices.","authors":"Rich Caruana, Radu S Niculescu, R Bharat Rao, Cynthia Simms","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>We apply machine learning to the problem of subpopulation assessment for Caesarian Section. In subpopulation assessment, we are interested in making predictions not for a single patient, but for groups of patients. Typically, in any large population, different subpopulations will have different \"outcome\" rates. In our example, the C-section rate of a population of 22,176 expectant mothers is 16.8%; yet, the 17 physician groups that serve this population have vastly different group C-section rates, ranging from 11% to 23%. The ultimate goal of subpopulation assessment is to determine if these variations in the observed rates can be attributed to (a) variations in intrinsic risk of the patient sub-populations (i.e. some groups contain more \"high-risk C-section\" patients), or (b) differences in physician practice (i.e. some groups do more C-sections). Our results indicate that although there is some variation in intrinsic risk, there is also much variation in physician practice.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244521/pdf/procamiasymp00001-0167.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22140173","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Biomedical informatics in general and pharmacogenomics in particular require a research platform that simultaneously enables discovery while protecting research subjects' privacy and information confidentiality. The development of inexpensive DNA sequencing and analysis technologies promises unprecedented database access to very specific information about individuals. To allow analysis of this data without compromising the research subjects' privacy, we must develop methods for removing identifying information from medical and genomic data. In this paper, we build upon the idea that binned database records are more difficult to trace back to individuals. We represent symbolic and numeric data hierarchically, and bin them by generalizing the records. We measure the information loss due to binning using an information theoretic measure called mutual information. The results show that we can bin the data to different levels of precision and use the bin size to control the tradeoff between privacy and data resolution.
{"title":"Using binning to maintain confidentiality of medical data.","authors":"Zhen Lin, Michael Hewett, Russ B Altman","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>Biomedical informatics in general and pharmacogenomics in particular require a research platform that simultaneously enables discovery while protecting research subjects' privacy and information confidentiality. The development of inexpensive DNA sequencing and analysis technologies promises unprecedented database access to very specific information about individuals. To allow analysis of this data without compromising the research subjects' privacy, we must develop methods for removing identifying information from medical and genomic data. In this paper, we build upon the idea that binned database records are more difficult to trace back to individuals. We represent symbolic and numeric data hierarchically, and bin them by generalizing the records. We measure the information loss due to binning using an information theoretic measure called mutual information. The results show that we can bin the data to different levels of precision and use the bin size to control the tradeoff between privacy and data resolution.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244360/pdf/procamiasymp00001-0495.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22140182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective: To investigate how training older adults to find medical information using the Internet affects their locus of control.
Methods: Quantitative methods were utilized. Specifically, the Multidimensional Health Locus of Control survey was distributed at the onset of each seminar and again at the conclusion.
Results: Paired t-tests revealed that the subjects did not change their locus of control regarding their health beliefs over the period of the seminar. However, there was statistical significance with regard to eight specific questions.
Conclusion: Subjects scored high on their level of internal locus of control coming into the study. The majority of subjects had already learned to use the computer, owned a home computer, and had access to the Internet, but had not used the Internet to search for healthcare information. The challenge continues to be reaching those older adults who have not encountered the computer and the Internet.
{"title":"The internet and locus of control in older adults.","authors":"Robert J Campbell, Kimberly D Harris, James Wabby","doi":"","DOIUrl":"","url":null,"abstract":"<p><strong>Objective: </strong>To investigate how training older adults to find medical information using the Internet affects their locus of control.</p><p><strong>Methods: </strong>Quantitative methods were utilized. Specifically, the Multidimensional Health Locus of Control survey was distributed at the onset of each seminar and again at the conclusion.</p><p><strong>Results: </strong>Paired t-tests revealed that the subjects did not change their locus of control regarding their health beliefs over the period of the seminar. However, there was statistical significance with regard to eight specific questions.</p><p><strong>Conclusion: </strong>Subjects scored high on their level of internal locus of control coming into the study. The majority of subjects had already learned to use the computer, owned a home computer, and had access to the Internet, but had not used the Internet to search for healthcare information. The challenge continues to be reaching those older adults who have not encountered the computer and the Internet.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244513/pdf/procamiasymp00001-0137.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22140249","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Patrica J Tidmarsh, Joseph Cummings, William R Hersh, Charles P Freidman
The curricula of most medical informatics training programs are incomplete. We used Internet2-based videoconferencing to expand the educational opportunities of medical informatics students at Oregon Health & Science University and the University of Pittsburgh. Students and faculty in both programs shared extra-curricular research conferences and journal club meetings. A course in Information Retrieval was made available to students in both programs. The conferences, meetings and class were well accepted by participants. A few problems were experienced with the technology, some of which were resolved, and some non-technical challenges to distributing academic conferences, meetings and coursework were also uncovered. We plan to continue our efforts with expanded course and extra-curricular offerings and a more comprehensive evaluation strategy.
{"title":"Distributed medical informatics education using internet2.","authors":"Patrica J Tidmarsh, Joseph Cummings, William R Hersh, Charles P Freidman","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>The curricula of most medical informatics training programs are incomplete. We used Internet2-based videoconferencing to expand the educational opportunities of medical informatics students at Oregon Health & Science University and the University of Pittsburgh. Students and faculty in both programs shared extra-curricular research conferences and journal club meetings. A course in Information Retrieval was made available to students in both programs. The conferences, meetings and class were well accepted by participants. A few problems were experienced with the technology, some of which were resolved, and some non-technical challenges to distributing academic conferences, meetings and coursework were also uncovered. We plan to continue our efforts with expanded course and extra-curricular offerings and a more comprehensive evaluation strategy.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244240/pdf/procamiasymp00001-0828.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22140716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jerome K Wang, M Michael Shabot, Raymond G Duncan, Jeanette X Polaschek, Douglas T Jones
Many of the benefits of computerized physician order entry (CPOE) stem from its ability to support medical decision-making and error-reduction during patient care. This automated "intelligence" is typically represented by a network of rules. We describe a taxonomic representation of clinical decision-support rules in the context of developing and implementing a de novo CPOE and decision-support system. In our experience, this clinical rules taxonomy facilitated our implementation goals in the areas of physician acceptance and approval, rules construction and maintenance, and technical development and testing. This rules taxonomy may eventually be used to establish standards by which CPOE-based decision-support is measured.
{"title":"A clinical rules taxonomy for the implementation of a computerized physician order entry (CPOE) system.","authors":"Jerome K Wang, M Michael Shabot, Raymond G Duncan, Jeanette X Polaschek, Douglas T Jones","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>Many of the benefits of computerized physician order entry (CPOE) stem from its ability to support medical decision-making and error-reduction during patient care. This automated \"intelligence\" is typically represented by a network of rules. We describe a taxonomic representation of clinical decision-support rules in the context of developing and implementing a de novo CPOE and decision-support system. In our experience, this clinical rules taxonomy facilitated our implementation goals in the areas of physician acceptance and approval, rules construction and maintenance, and technical development and testing. This rules taxonomy may eventually be used to establish standards by which CPOE-based decision-support is measured.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244545/pdf/procamiasymp00001-0901.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22140723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The task of scheduling medical staff for evening rounds in the Clinical Teaching Unit of the Ottawa Hospital is a long complicated task due to its complexity. Three main classifications of staff, combined with various rotations, skill sets, clinical teams and vacation periods have combined to create a difficult scheduling problem. As there were no commercial packages available to solve this particular task, a study was made of heuristic scheduling and optimization techniques and a program based on a variation of the tabu search heuristic was written and tested. This system is being used to schedule medical staff at the Ottawa Hospital.
{"title":"Optimization of clinical teaching unit call schedules at the Ottawa hospital through tabu search heuristics.","authors":"Christine A White, George M White","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>The task of scheduling medical staff for evening rounds in the Clinical Teaching Unit of the Ottawa Hospital is a long complicated task due to its complexity. Three main classifications of staff, combined with various rotations, skill sets, clinical teams and vacation periods have combined to create a difficult scheduling problem. As there were no commercial packages available to solve this particular task, a study was made of heuristic scheduling and optimization techniques and a program based on a variation of the tabu search heuristic was written and tested. This system is being used to schedule medical staff at the Ottawa Hospital.</p>","PeriodicalId":79712,"journal":{"name":"Proceedings. AMIA Symposium","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2002-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2244436/pdf/procamiasymp00001-0930.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"22140729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}