Pub Date : 2018-10-01DOI: 10.1109/EMBC.2016.7590722
F. Nougarou, Alexandre Campeau-Lecours, R. Islam, Daniel Massicotte, Benoit Gosselin
An efficient pattern recognition system based exclusively on forearm surface Electromyographic (sEMG) signals is proposed to provide a more intuitive control of a robotic arm used by some of the disabled. The main contribution of this paper is the use of an original set of features characterizing the muscle activity distribution obtained with high-density sEMG (HD sEMG) sensors. Contrary to simple sEMG, HD sEMG can produce muscle activity images with spatial distributions that differ according to forearm movement. In order to translate this distribution, the proposed set of features includes the center of gravity, the mean amplitude and the percentage of influence computed in each HD sEMG image divided in sub-images. Based on these features, the recognition system locates nine forearm movements with high classification accuracies (99.23%). The results in terms of the number of learning data, the image resolutions (spatial filtering) and the number of sub-images demonstrate the potential of the proposed recognition system and its good performance-complexity trade-off.
{"title":"Muscle Activity Distribution Features Extracted from HD sEMG to Perform Forearm Pattern Recognition","authors":"F. Nougarou, Alexandre Campeau-Lecours, R. Islam, Daniel Massicotte, Benoit Gosselin","doi":"10.1109/EMBC.2016.7590722","DOIUrl":"https://doi.org/10.1109/EMBC.2016.7590722","url":null,"abstract":"An efficient pattern recognition system based exclusively on forearm surface Electromyographic (sEMG) signals is proposed to provide a more intuitive control of a robotic arm used by some of the disabled. The main contribution of this paper is the use of an original set of features characterizing the muscle activity distribution obtained with high-density sEMG (HD sEMG) sensors. Contrary to simple sEMG, HD sEMG can produce muscle activity images with spatial distributions that differ according to forearm movement. In order to translate this distribution, the proposed set of features includes the center of gravity, the mean amplitude and the percentage of influence computed in each HD sEMG image divided in sub-images. Based on these features, the recognition system locates nine forearm movements with high classification accuracies (99.23%). The results in terms of the number of learning data, the image resolutions (spatial filtering) and the number of sub-images demonstrate the potential of the proposed recognition system and its good performance-complexity trade-off.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115031133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572122
L. Lazli, M. Boukadoum, O. Ait Mohamed
We describe a computer-aided diagnosis (CAD) system for discriminating patients suffering from Alzheimer's disease (AD) dementia and healthy patients. It is based on: 1) a clustering process to assess white matter, gray matter and cerebrospinal fluid volumes from noisy anatomical magnetic resonance (MR) and functional positron emission tomography (PET) brain images11The MR and PET data used in this work were obtained from the Alzheimer's disease Neuroimaging Initiative (ADNI) database (http://adni.loni.usc.edu/).; 2) a classification process that distinguishes the brain images of normal and AD patients. The clustering stage consists of three steps: First, the fuzzy c-means (FCM) algorithm is used for a fuzzy partition of the initial class centroids. Second, fuzzy tissue maps are computed using a possibilistic C-means (PCM) algorithm that uses the FCM partition to obtain the final image clusters. The final segmentation is then made to delimit the brain tissue volumes. For the classification stage, a support vector machine (SVM) is used with different kernel functions. Validating the proposed CAD system on the MRI and PET images of 45 AD and 50 healthy brains, of subjects aged between 55 and 90 years, shows better sensitivity, specificity and accuracy in comparison to three alternative approaches, namely FCM, PCM and VAF (Voxels-As-Features). The accuracy rates for the noisiest images (20% of noise) were 75% for MRI and 73% for PET scan, compared to 71 % and 70,2%, 68.5% and 67%, and 65 % and 64.7 % with the three other approaches, respectively.
{"title":"Computer-Aided Diagnosis System for Alzheimer's Disease Using Fuzzy-Possibilistic Tissue Segmentation and SVM Classification","authors":"L. Lazli, M. Boukadoum, O. Ait Mohamed","doi":"10.1109/LSC.2018.8572122","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572122","url":null,"abstract":"We describe a computer-aided diagnosis (CAD) system for discriminating patients suffering from Alzheimer's disease (AD) dementia and healthy patients. It is based on: 1) a clustering process to assess white matter, gray matter and cerebrospinal fluid volumes from noisy anatomical magnetic resonance (MR) and functional positron emission tomography (PET) brain images11The MR and PET data used in this work were obtained from the Alzheimer's disease Neuroimaging Initiative (ADNI) database (http://adni.loni.usc.edu/).; 2) a classification process that distinguishes the brain images of normal and AD patients. The clustering stage consists of three steps: First, the fuzzy c-means (FCM) algorithm is used for a fuzzy partition of the initial class centroids. Second, fuzzy tissue maps are computed using a possibilistic C-means (PCM) algorithm that uses the FCM partition to obtain the final image clusters. The final segmentation is then made to delimit the brain tissue volumes. For the classification stage, a support vector machine (SVM) is used with different kernel functions. Validating the proposed CAD system on the MRI and PET images of 45 AD and 50 healthy brains, of subjects aged between 55 and 90 years, shows better sensitivity, specificity and accuracy in comparison to three alternative approaches, namely FCM, PCM and VAF (Voxels-As-Features). The accuracy rates for the noisiest images (20% of noise) were 75% for MRI and 73% for PET scan, compared to 71 % and 70,2%, 68.5% and 67%, and 65 % and 64.7 % with the three other approaches, respectively.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128523803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572092
McNiel-Inyani Keri, A. W. Shehata, Quinn A. Boser, A. Vette, Jacqueline S. Hebert
Limb amputation affects many individuals across the world, with the majority of amputations occurring in the lower limb. Healthy individuals with intact limbs have biological sensors embedded in their anatomy to interact with the environment and to facilitate stable walking. Lower limb prosthetic users lose these embedded sensors, leading to decreased balance and an increased risk of falling, abnormal gait, and decreased quality of life. Tactile and kinesthetic sensory feedback techniques are being investigated for upper limb prosthetic users and may soon translate to lower limb users. A barrier to implementing these techniques is the lack of adequate instrumentation of lower limb prostheses. The objective of this research was to design and develop a low-cost wireless system, using inertial measurement units, which can detect when a single axis prosthetic knee is in motion. This sensor could be used to communicate the movement of a prosthetic device to actuators responsible for providing feedback to the user. Our results indicate that the device is capable of tracking the onset and termination of movement at normal walking speeds.
{"title":"Development and Verification of a Low-Cost Prosthetic Knee Motion Sensor","authors":"McNiel-Inyani Keri, A. W. Shehata, Quinn A. Boser, A. Vette, Jacqueline S. Hebert","doi":"10.1109/LSC.2018.8572092","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572092","url":null,"abstract":"Limb amputation affects many individuals across the world, with the majority of amputations occurring in the lower limb. Healthy individuals with intact limbs have biological sensors embedded in their anatomy to interact with the environment and to facilitate stable walking. Lower limb prosthetic users lose these embedded sensors, leading to decreased balance and an increased risk of falling, abnormal gait, and decreased quality of life. Tactile and kinesthetic sensory feedback techniques are being investigated for upper limb prosthetic users and may soon translate to lower limb users. A barrier to implementing these techniques is the lack of adequate instrumentation of lower limb prostheses. The objective of this research was to design and develop a low-cost wireless system, using inertial measurement units, which can detect when a single axis prosthetic knee is in motion. This sensor could be used to communicate the movement of a prosthetic device to actuators responsible for providing feedback to the user. Our results indicate that the device is capable of tracking the onset and termination of movement at normal walking speeds.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124333215","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572257
Paul E. Stevenson, J. Christen
The growing market for wearable, portable, and IoT devices has generated a need for a class of circuits to meet the requirements for these applications. In this work we specifically investigate ammeters. The design space requires low component count circuits for measuring slowly varying currents using low-cost microcontrollers. Simple architectures, feasible for an electronics novice are described and compared experimentally. The use of the time domain to improve error and range of measurement is considered. This guide provides an individual without extensive electronics design experience with a simple selection guide for choosing the appropriate architecture for their specific application.
{"title":"A Practical Guide to Circuit Selection for Portable Microprocessor-Based, Low Component Count, Near-DC Ammeter for Custom Instruments","authors":"Paul E. Stevenson, J. Christen","doi":"10.1109/LSC.2018.8572257","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572257","url":null,"abstract":"The growing market for wearable, portable, and IoT devices has generated a need for a class of circuits to meet the requirements for these applications. In this work we specifically investigate ammeters. The design space requires low component count circuits for measuring slowly varying currents using low-cost microcontrollers. Simple architectures, feasible for an electronics novice are described and compared experimentally. The use of the time domain to improve error and range of measurement is considered. This guide provides an individual without extensive electronics design experience with a simple selection guide for choosing the appropriate architecture for their specific application.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124132359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572227
S. Gill, Suraj Nssk, N. Seth, E. Scheme
Increases in the rates of chronic disease and an aging population have created a demand for new forms of preventative care and proactive health monitoring technologies. While senior populations may be hesitant to adopt wearable technologies, the ability to retrofit assistive devices already in use by the individuals may provide a major stepping stone for increased adoption rates and monitoring abilities. Design of such systems often exhibit challenges with respect to sensor selection, placement, and consequently, reliability and usability of the system in real-world environments. As part of a growing line of smart assistive devices, this work presents a proposed design for a multi-sensor walker with pilot data collected and tested in a real-world environment, including outdoors. Preliminary analysis of results demonstrates the ability to determine levels of activity and environments, important factors related to health and wellness and risk of falls.
{"title":"Design of a Smart IoT-Enabled Walker for Deployable Activity and Gait Monitoring","authors":"S. Gill, Suraj Nssk, N. Seth, E. Scheme","doi":"10.1109/LSC.2018.8572227","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572227","url":null,"abstract":"Increases in the rates of chronic disease and an aging population have created a demand for new forms of preventative care and proactive health monitoring technologies. While senior populations may be hesitant to adopt wearable technologies, the ability to retrofit assistive devices already in use by the individuals may provide a major stepping stone for increased adoption rates and monitoring abilities. Design of such systems often exhibit challenges with respect to sensor selection, placement, and consequently, reliability and usability of the system in real-world environments. As part of a growing line of smart assistive devices, this work presents a proposed design for a multi-sensor walker with pilot data collected and tested in a real-world environment, including outdoors. Preliminary analysis of results demonstrates the ability to determine levels of activity and environments, important factors related to health and wellness and risk of falls.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"348 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132310694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572041
C. Castagneri, V. Agostini, G. Balestra, M. Knaflitz, M. Carlone, Giuseppe Massazza
The study of EMG cycle patterns is an important tool in clinical research, for managing locomotion pathologies and rehabilitation. Statistical Gait Analysis (SGA) was introduced to process muscle cyclic activation patterns extracted from a functional walk. The CIMAP algorithm was recently introduced to improve the SGA. As result of CIMAP, principal activations, defined as those activations necessary to perform a specific cyclic movement, are extracted. They are coded using a binary string of activation values that characterizes a specific muscle. The aim of this work is to define an index to evaluate muscle-activation asymmetry in cyclic movements, using principal activations. The index was significantly higher in patients with knee megaprosthesis, with respect to healthy controls, for tibialis anterior, rectus femoris and lateral hamstring.
{"title":"EMG Asymmetry Index in Cyclic Movements","authors":"C. Castagneri, V. Agostini, G. Balestra, M. Knaflitz, M. Carlone, Giuseppe Massazza","doi":"10.1109/LSC.2018.8572041","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572041","url":null,"abstract":"The study of EMG cycle patterns is an important tool in clinical research, for managing locomotion pathologies and rehabilitation. Statistical Gait Analysis (SGA) was introduced to process muscle cyclic activation patterns extracted from a functional walk. The CIMAP algorithm was recently introduced to improve the SGA. As result of CIMAP, principal activations, defined as those activations necessary to perform a specific cyclic movement, are extracted. They are coded using a binary string of activation values that characterizes a specific muscle. The aim of this work is to define an index to evaluate muscle-activation asymmetry in cyclic movements, using principal activations. The index was significantly higher in patients with knee megaprosthesis, with respect to healthy controls, for tibialis anterior, rectus femoris and lateral hamstring.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128887436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572083
Tamer AbdEIFatah, M. Jalali, S. Mahshid
Here we report on design, fabrication and implementation of a nanosurfac microfluidic device for efficient bacteria capture and optical detection. The device features simple design and ease of implementation. The principal of operation depends on the self-assembly of microparticles (polystyrene particles) at a pillar array region to form a Nano-filter for subsequent bacteria capture on gold nano/micro islands. The design was optimized using 2D COMSOL simulation. The device was fabricated using a single UV lithography step followed by electrodeposition of the gold structures and a subsequent step of polydimethylsiloxane (PDMS) bonding for device sealing. Lastly, the device was experimentally implemented using Escherichia coli (E. coli) bacteria showing efficient bacteria capturing performance.
{"title":"A Nanosurface Microfluidic Device for Capture and Detection of Bacteria","authors":"Tamer AbdEIFatah, M. Jalali, S. Mahshid","doi":"10.1109/LSC.2018.8572083","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572083","url":null,"abstract":"Here we report on design, fabrication and implementation of a nanosurfac microfluidic device for efficient bacteria capture and optical detection. The device features simple design and ease of implementation. The principal of operation depends on the self-assembly of microparticles (polystyrene particles) at a pillar array region to form a Nano-filter for subsequent bacteria capture on gold nano/micro islands. The design was optimized using 2D COMSOL simulation. The device was fabricated using a single UV lithography step followed by electrodeposition of the gold structures and a subsequent step of polydimethylsiloxane (PDMS) bonding for device sealing. Lastly, the device was experimentally implemented using Escherichia coli (E. coli) bacteria showing efficient bacteria capturing performance.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134487229","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572113
S. Candemir, S. Rajaraman, G. Thoma, Sameer Kiran Antani
This study investigates using deep convolutional neural networks (CNN) for automatic detection of cardiomegaly in digital chest X-rays (CXRs). First, we employ and fine-tune several deep CNN architectures to detect presence of cardiomegaly in CXRs. Next, we introduce a CXR-based pre-trained model where we first fully train an architecture with a very large CXR dataset and then fine-tune the system with cardiomegaly CXRs. Finally, we investigate the correlation between softmax probability of an architecture and the severity of the disease. We use two publicly available datasets, NLM-Indiana Collection and NIH-CXR datasets. Based on our preliminary results (i) data-driven approach produces better results than prior rule-based approaches developed for cardiomegaly detection, (ii) our preliminary experiment with alternative pre-trained model is promising, and (iii) the system is more confident if severity increases.
{"title":"Deep Learning for Grading Cardiomegaly Severity in Chest X-Rays: An Investigation","authors":"S. Candemir, S. Rajaraman, G. Thoma, Sameer Kiran Antani","doi":"10.1109/LSC.2018.8572113","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572113","url":null,"abstract":"This study investigates using deep convolutional neural networks (CNN) for automatic detection of cardiomegaly in digital chest X-rays (CXRs). First, we employ and fine-tune several deep CNN architectures to detect presence of cardiomegaly in CXRs. Next, we introduce a CXR-based pre-trained model where we first fully train an architecture with a very large CXR dataset and then fine-tune the system with cardiomegaly CXRs. Finally, we investigate the correlation between softmax probability of an architecture and the severity of the disease. We use two publicly available datasets, NLM-Indiana Collection and NIH-CXR datasets. Based on our preliminary results (i) data-driven approach produces better results than prior rule-based approaches developed for cardiomegaly detection, (ii) our preliminary experiment with alternative pre-trained model is promising, and (iii) the system is more confident if severity increases.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127094712","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572286
A. Vetek, Kiti Müller, H. Lindholm
Sleep stage classification is usually performed by trained professionals using visual inspection of bio-electrical recordings from a subject and is the first step in quantifying the quality of sleep and diagnosing sleep disorders. We introduce an extensible, modality-agnostic deep learning system to automate the task of temporal sleep stage classification from raw electroencephalography, electrooculography and electromyography signals. The proposed architecture uses a combination of convolutional neural networks (CNN) and recurrent neural networks (RNN). The compact size of the system makes it not only computationally efficient but also more appropriate for smaller datasets. We evaluated the proposed system on a sleep dataset collected in a home environment from healthy subjects and found that the incorporation of temporal information (sleep stage transitions) boosted overall performance in terms of macro-average F1 scores, and in particular provided a significant improvement for the worst performing class, N1 compared to other approaches.
{"title":"A Compact Deep Learning Network for Temporal Sleep Stage Classification","authors":"A. Vetek, Kiti Müller, H. Lindholm","doi":"10.1109/LSC.2018.8572286","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572286","url":null,"abstract":"Sleep stage classification is usually performed by trained professionals using visual inspection of bio-electrical recordings from a subject and is the first step in quantifying the quality of sleep and diagnosing sleep disorders. We introduce an extensible, modality-agnostic deep learning system to automate the task of temporal sleep stage classification from raw electroencephalography, electrooculography and electromyography signals. The proposed architecture uses a combination of convolutional neural networks (CNN) and recurrent neural networks (RNN). The compact size of the system makes it not only computationally efficient but also more appropriate for smaller datasets. We evaluated the proposed system on a sleep dataset collected in a home environment from healthy subjects and found that the incorporation of temporal information (sleep stage transitions) boosted overall performance in terms of macro-average F1 scores, and in particular provided a significant improvement for the worst performing class, N1 compared to other approaches.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127347462","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-10-01DOI: 10.1109/LSC.2018.8572194
S. Rosati, C. M. Gianfreda, G. Balestra, V. Giannini, S. Mazzetti, D. Regge
According to the guidelines, patients with locally advanced colorectal cancer undergo neoadjuvant chemotherapy. However, response to therapy is reached only up to 30% of cases. Therefore, it would be important to predict response to therapy before treatment. In this study, we demonstrated that the simultaneous optimization of feature subset and classifier parameters on different imaging datasets (T2w, DWI and PET) could improve classification performance. On a dataset of 51 patients (21 responders, 30 non responders), we obtained an accuracy of 90%, 84% and 76% using three optimized SVM classifiers fed with selected features from PET, T2w and ADC images, respectively.
{"title":"Radiomics to Predict Response to Neoadjuvant Chemotherapy in Rectal Cancer: Influence of Simultaneous Feature Selection and Classifier Optimization","authors":"S. Rosati, C. M. Gianfreda, G. Balestra, V. Giannini, S. Mazzetti, D. Regge","doi":"10.1109/LSC.2018.8572194","DOIUrl":"https://doi.org/10.1109/LSC.2018.8572194","url":null,"abstract":"According to the guidelines, patients with locally advanced colorectal cancer undergo neoadjuvant chemotherapy. However, response to therapy is reached only up to 30% of cases. Therefore, it would be important to predict response to therapy before treatment. In this study, we demonstrated that the simultaneous optimization of feature subset and classifier parameters on different imaging datasets (T2w, DWI and PET) could improve classification performance. On a dataset of 51 patients (21 responders, 30 non responders), we obtained an accuracy of 90%, 84% and 76% using three optimized SVM classifiers fed with selected features from PET, T2w and ADC images, respectively.","PeriodicalId":254835,"journal":{"name":"2018 IEEE Life Sciences Conference (LSC)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130864102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：