Journal of Medical and Biological Engineering最新文献

Purpose: Alzheimer's disease (AD), a neurodegenerative disorder, is a condition that impairs cognition, memory, and behavior. Mild cognitive impairment (MCI), a transitional stage before AD, urgently needs the development of prediction models for conversion from MCI to AD.

Method: This study used machine learning methods to predict whether MCI subjects would develop AD, highlighting the importance of biomarkers (biological indicators from neuroimaging, such as MRI and PET scans, and molecular assays from cerebrospinal fluid or blood) and non-biomarker features in AD research and clinical practice. These indicators aid in early diagnosis, disease monitoring, and the development of potential treatments for MCI subjects. Using baseline data, which includes measurements of different biomarkers, we predicted disease progression at the patient's last visit. The Shapley value explanation (SHAP) technique was used to identify key features for predicting patient progression.

Results: The study used the ADNI database to evaluate the effectiveness of eight classification methods for predicting progression from MCI to AD. Four fundamental data sampling approaches were compared to balance the dataset and reduce overfitting. The SHAP technique improved the ability to identify biomarkers and non-biomarker features, enhancing the prediction of disease progression. NEAR-MISS was found to be the most advantageous sampling method, while XGBoost was found to be the superior classification method, offering enhanced accuracy and predictive power.

Conclusion: The proposed SHAP for feature selection combined with XGBoost may provide improved predictive accuracy in diagnosing Alzheimer's patients.

Purpose

In this study, animal experiments and finite element simulation was used to study pronation-supination and the mechanical environment of the elbow at different stages of joint capsule healing after injury to reveal the mechanism of mechanical factors affecting the development of the elbow contracture.

Methods

The injured limbs were immobilized for 2 weeks, 4 weeks, 6 weeks and 8 weeks in animal experiments. The total ROM and joint stiffness of the limbs were tested at each time point to define the differences between different groups. The internal soft tissue stress during joint capsule healing was obtained by finite element simulation.

Results

In the animal experiments, compared with the no injury group, the total ROM of the injury group was decreased and the joint stiffness of the injury group was increased. The joint capsule thickness and shear modulus increased gradually in different stages of joint capsule healing. In the simulation, the internal soft tissue stress increased at different stages of joint capsule healing. The peak stress of the capsule was 8.32 MPa, 10.5 MPa, 11.4 MPa, 11.1 MPa and 12.2 MPa, respectively. The peak stress of humerus cartilage at different stages of joint capsule healing was greater than that of ulna cartilage and 1.2 times that of ulna cartilage.

Conclusion

This study revealed the evolution of stress distribution in the surrounding tissue in the process of joint capsule healing and provided a theoretical basis for the rehabilitation of the elbow contracture and the design of rehabilitation aids.

Purpose

Lung cancer is the leading cause of cancer-related death. Early detection and treatment are crucial to improve survival rates. Radiologists determine whether the nodules are benign or malignant by observing their morphological attributes. However, this can be a challenging task for well-trained doctors.

Methods

We propose a more efficient automatic lung nodule analysis method, which establishes a clear cause-and-effect logic relationship between attribute features and malignancy features by incorporating multiple instance learning (MIL). The designed MIL classifier aggregates the learned instance weights and corresponding attribute features to form malignancy features. Compared to existing methods, it starts by mirroring the way radiologists observe nodules, then proceeds to extract the multi-scale morphological attribute characteristics of the nodules. The instance weight also serves as the attribute score of the attribute, providing a reference for consultation.

Results

Our method was validated using the LIDC-IDRI dataset and achieved an accuracy of 93.05% on benign-malignant classification task with the added capability of accurately scoring the attributes.

Conclusion

The proposed method based on attribute score regression and multi-instance learning establishes the causal relationship between attribute scores and malignancy. This method improves accuracy in nodule classification and addresses the issue of poor model interpretability.

Purpose

The purpose of this study was to investigate the effects of implant spacing and placement angle on peri-implant bone stress using the finite element method.

Methods

A model of the maxilla of an edentulous patient was obtained by computed tomography, and a splint prosthesis consisting of short or tilted implants was applied to the maxillary posterior region. Three spacings (15, 17, and 19 mm) were used for each restoration, and three placement angles (− 5, 0, and 5 degrees) were available at each spacing. Finite element analysis was used to evaluate the stress distributions and resonant frequencies of all the models by applying both vertical and oblique forces to the splint prostheses simultaneously.

Results

The patters of stress distribution were better in the short implant group (19 mm with a − 5 degree placement angle) and the tilted implant group (19 mm with a 5 degree placement angle). In the short implant group, the difference in cortical bone stress (i.e., the difference between the maximum and minimum principal stresses) decreased from 248 MPa (15 mm, 5 degrees) to 116 MPa (19 mm, − 5 degrees), and the corresponding maximum von Mises stress in the implants was reduced by 35%. The resonance frequency of the short implant group (5300–5500 Hz) was slightly lower than that of the tilted implant group (5400–5700 Hz).

Conclusion

Implant spacing and the placement angle significantly affect the peri-implant bone stress distribution, proper implant placement is essential to minimize this risk.

Purpose

Owing to the increased mortality of heart diseases worldwide, especially myocardial infarction (MI), early detection is essential for improved diagnosis and treatment. The main purpose of this study is to develop a myocardial infarction detection method that combines deep learning and image processing, focusing on abnormalities in left ventricular (LV) wall motion.

Methods

The proposed method primarily uses the LV wall motion movement as a feature to train an LSTM network for MI detection. LV wall motion annotated by expert cardiologists was used as the ground truth. Accuracy, sensitivity, specificity, and area under the curve (AUC) were used to evaluate model performance. The proposed method primarily uses LV wall motion as a feature, combined with LV size and image pixels, to improve diagnostic accuracy over existing computer-aided design (CAD) systems.

Results

The LSTM model achieved the highest diagnostic performance when trained on a combination of LV wall motion, LV size, and image pixel features with an accuracy of 95%, sensitivity of 96%, specificity of 94%, and an AUC value of 0.98. The LSTM model significantly outperformed models trained on individual feature sets or conventional machine learning algorithms. The inclusion of LV wall motion analysis improved accuracy by 10% compared to using only LV size and pixel data.

Conclusion

Our MI diagnosis system uses echocardiographic image analysis and LSTM-based deep learning to accurately detect LV wall motion issues related to MI. Compared with current CAD systems, the inclusion of LV wall motion analysis significantly improves diagnosis accuracy. The proposed system can help physicians detect MI early, thereby accelerating treatment and improving patient outcomes.

Purpose

Small animals are integral to medical research as they provide insights into human diseases. Mice, particularly suitable as human models, share gene functions, making them vital for biomedical research. Positron emission tomography (PET) has emerged as a key tool for non-invasive imaging of mouse models, providing molecular-level insights with remarkable sensitivity. Achieving optimal spatial resolution is crucial for capturing detailed images of small animal organs, and enhancing sensitivity is imperative for microPET scanner efficiency.

Methods

This study investigates the performance of microPET scanners using cuboidal and tapered arrays, simulated by the GATE Monte Carlo package. To this end, it focuses on detector materials, crystal geometry, and reconstruction methods. Finally, critical parameters such as sensitivity, NECR, and FWHM of Gaussian fit of image intensity profiles are assessed.

Results

Simulation outputs reveal that tapered arrays outperform their cuboid counterparts by 44% in sensitivity and NECR, along with a 22% improvement in spatial resolution. The relative FWHM difference for crystals compared to LSO remains below 5%. Crystal material significantly affects sensitivity and NECR, with BGO demonstrating 25% greater values than LSO. Meanwhile, GSO and LYSO showed 32% and 60% lower values, respectively. BGO crystal demonstrated a higher profile amplitude, indicating higher counts. This difference could be attributed more to heightened noise than an increase in signal, as BGO crystals exhibit a higher scatter fraction than other crystals. Furthermore, COSEM and ACOSEM algorithms achieve the minimum FWHM of 0.7 mm, suggesting 10% better spatial resolution than the OSEM algorithm. In contrast, RAMLA and MRAMLA algorithms showed 14% and 4% worse spatial resolution than the OSEM algorithm, respectively.

Conclusion

Tapered arrays, especially when paired with BGO crystals, demonstrate superior sensitivity, NECR, and lower FWHM suggesting better spatial resolution than cuboid arrays. Crystal material choice minimally affects FWHM for a low-activity point source but significantly influences sensitivity and NECR, with BGO outperforming other crystals. COSEM and ACOSEM reconstruction algorithms yielded better image quality with lower FWHM and noise, demonstrating their effectiveness in microPET applications.

Purpose

Radiotherapy (RT) is a commonly employed therapeutic strategy for the treatment of localized cancers, including prostate cancer (PCa). Despite significant advancements in radiotherapy technology over recent years, high recurrence and metastasis of PCa after RT remain critical challenges. Various mechanisms have been implicated in how cancer evades radiotherapy, and exosomes, a type of extracellular vesicles (EVs) has recently been identified as one of the contributing factors. This study aimed to investigate whether exosomes derived from irradiated PCa cells are involved in the cancer progression and to identify possible key factor in this process.

Methods

Exosomes were isolated from irradiated or non-irradiated PCa cell lines (designated as Rad-Exo or Exo) and characterized by specific marker expression, morphology and size. PCa cells treated with Rad-Exo or Exo were analyzed for the effects of proliferation, specific gene expression, migration and cancer stem cell property. Differential protein expression in Rad-Exo and Exo were carried out by mass spectrometry.

Results

Results showed that, compared to Exo, Rad-Exo treatment inhibited cell proliferation but significantly promoted migration and elevated the expression of genes related to epithelial to mesenchymal transition. Additionally, cells treated with Rad-Exo showed increased expression of genes related to cancer stem cells. Mass spectrometry identified POTEE as more abundant within Rad-Exo then in Exo, and its expression was confirmed to be elevated in PCa cells following irradiation. Furthermore, POTEE expression increased in cells after Rad-Exo treatment.

Conclusion

This study suggests that exosomes derived from irradiated PCa cells may function as a driver of cancer progression, including recurrent or metastatic cancer. Also, exosomal POTEE may serve as a potential target for future therapeutic or diagnostic investigations in prostate cancer.

Purpose

To assess if transfer learning strategies can improve stroke patients’ ability to control a Brain-computer interface (BCI) based on motor intention across an upper extremity neurorehabilitation intervention.

Methods

Three subject-specific session-to-session training strategies were retrospectively assessed in the present study, using information acquired during a BCI intervention in 12 stroke patients. One strategy used data from the previous therapy session (previous session), another used data from all previous sessions (accumulative) and another initially used previous session’s data and was updated with data acquired during the current session (instantaneous).

Results

Classification accuracy was significantly higher with the instantaneous strategy (median = 76.4%, IQR = [68.7%, 81.5%]) compared to the obtained with the accumulative (71.67%, [65.1%, 78.5%]) and previous session (69.2%, [59%, 77.4%]) strategies. Median classification accuracies across sessions were also higher with the instantaneous strategy in each BCI intervention session.

Conclusion

The instantaneous strategy could allow stroke patients to achieve a competitive level of BCI performance during a motor intention BCI intervention without reducing effective therapy time or requiring data from other patients.

Purpose

High-intensity focused ultrasound (HIFU) is a promising non-invasive technique for thermal ablation of tumors. Positioning the focal point of the HIFU accurately prior to the procedure is crucial to the success of the treatment. A change in backscattered energy (CBE) in ultrasound images has been shown to allow visualization of thermal information and can be used to locate the focal spot of HIFU prior to ablation. In CBE imaging, however, tailing artifacts may exist below the focal point of HIFU to hinder the identification of the HIFU focal spot.

Methods

This study proposed ultrasound delta CBE (DCBE) imaging that reduces CBE artifacts by local energy subtraction between measured and the reference envelope images. Phantom experiments were performed for validation of the proposed method. A HIFU system operating at a frequency of 2.12 MHz was used to heat phantoms, which were imaged with a clinical ultrasound scanner equipped with a 3-MHz convex transducer for analysis of CBE and DCBE data.

Results

The results showed that the DCBE value increases monotonically with temperature (correlation coefficient = 0.90). Particularly, DCBE imaging can identify the HIFU focal spot, suppress tailing artifacts, and increase the contrast between the focal and artifact zones by 8 dB in comparison with conventional CBE imaging.

Conclusion

Based on this study, DCBE imaging may be an effective method of locating HIFU focal points through ultrasound backscattered energy.

Purpose

Accurate predictions of postoperative recurrence are essential for determining appropriate follow-up treatments after surgery, as patients with non-small-cell lung cancer (NSCLC) at the same clinical stage have different recurrence incidences. However, simple convolutional neural network (CNN)-based methods are limited when presented with tumors of various sizes. This study aims to predict two-year recurrence-free survival precisely in patients with tumors of various sizes on preoperative CT images using what is termed a Tumor-Centric Attention Network (TCA-Net).

Methods

The proposed network features dual branches, each with an identical architecture but distinct weights to extract diverse features from CT images and tumor masks simultaneously. The tumor-centric attention module integrates two disparate feature maps at each level to amplify the characteristics of the tumor. All feature maps are concatenated with the finest resolution, enabling the extraction and integration of comprehensive multi-scale features for the complex tumor environment.

Results

TCA-Net showed an accuracy of 75%, balanced accuracy of 75.05%, specificity of 76.16% and an AUC value of 0.78. These results represent more balanced accuracies by 4.76% and 2.58% compared to ResNet-18 with CT images and dual ResNet-18s with CT images and tumor masks, respectively. Specifically, TCA-Net demonstrated a substantial improvement in the small-sized tumor group, achieving a balanced accuracy of 81.32%, sensitivity of 85.71%, and specificity of 76.92%.

Conclusion

TCA-Net improved the prediction performance of two-year recurrence-free survival on average across tumors of all sizes, with significant improvements, especially for small-sized tumors.