Deep learning-based prediction of indication for cracked tooth extraction using panoramic radiography.

BACKGROUND: We aimed to determine the feasibility of utilizing deep learning-based predictions of the indications for cracked tooth extraction using panoramic radiography. METHODS: Panoramic radiographs of 418 teeth (group 1: 209 normal teeth; group 2: 209 cracked teeth) were evaluated for the training and testing of a deep learning model. We evaluated the performance of the cracked diagnosis model for individual teeth using InceptionV3, ResNet50, and EfficientNetB0. The cracked tooth diagnosis model underwent fivefold cross-validation with 418 data instances divided into training, validation, and test sets at a ratio of 3:1:1. RESULTS: To evaluate the feasibility, the sensitivity, specificity, accuracy, and F1 score of the deep learning models were calculated, with values of 90.43-94.26%, 52.63-60.77%, 72.01-75.84%, and 76.36-79.00%, respectively. CONCLUSION: We found that the indications for cracked tooth extraction can be predicted to a certain extent through a deep learning model using panoramic radiography.

Airway and Airway Obstruction Site Segmentation Study Using U-Net with Drug-Induced Sleep Endoscopy Images.

Obstructive sleep apnea is characterized by a decrease or cessation of breathing due to repetitive closure of the upper airway during sleep, leading to a decrease in blood oxygen saturation. In this study, employing a U-Net model, we utilized drug-induced sleep endoscopy images to segment the major causes of airway obstruction, including the epiglottis, oropharynx lateral walls, and tongue base. The evaluation metrics included sensitivity, specificity, accuracy, and Dice score, with airway sensitivity at 0.93 (± 0.06), specificity at 0.96 (± 0.01), accuracy at 0.95 (± 0.01), and Dice score at 0.84 (± 0.03), indicating overall high performance. The results indicate the potential for artificial intelligence (AI)-driven automatic interpretation of sleep disorder diagnosis, with implications for standardizing medical procedures and improving healthcare services. The study suggests that advancements in AI technology hold promise for enhancing diagnostic accuracy and treatment efficacy in sleep and respiratory disorders, fostering competitiveness in the medical AI market.

Re-hospitalization factors and economic characteristics of urinary tract infected patients using machine learning.

Objective: Urinary tract infection is one of the most prevalent bacterial infectious diseases in outpatient treatment, and 50-80% of women experience it more than once, with a recurrence rate of 40-50% within a year; consequently, preventing re-hospitalization of patients is critical. However, in the field of urology, no research on the analysis of the re-hospitalization status for urinary tract infections using machine learning algorithms has been reported to date. Therefore, this study uses various machine learning algorithms to analyze the clinical and nonclinical factors related to patients who were re-hospitalized within 30 days of urinary tract infection. Methods: Data were collected from 497 patients re-hospitalized for urinary tract infections within 30 days and 496 patients who did not require re-hospitalization. The re-hospitalization factors were analyzed using four machine learning algorithms: gradient boosting classifier, random forest, naive Bayes, and logistic regression. Results: The best-performing gradient boosting classifier identified respiratory rate, days of hospitalization, albumin, diastolic blood pressure, blood urea nitrogen, body mass index, systolic blood pressure, body temperature, total bilirubin, and pulse as the top-10 factors that affect re-hospitalization because of urinary tract infections. The 993 patients whose data were collected were divided into risk groups based on these factors, and the re-hospitalization rate, days of hospitalization, and medical expenses were observed to decrease from the high- to low-risk group. Conclusions: This study showed new possibilities in analyzing the status of urinary tract infection-related re-hospitalization using machine learning. Identifying factors affecting re-hospitalization and incorporating preventable and reinforcement-based treatment programs can aid in reducing the re-hospitalization rate and average number of days of hospitalization, thereby reducing medical expenses.

Allergy Wheal and Erythema Segmentation Using Attention U-Net.

The skin prick test (SPT) is a key tool for identifying sensitized allergens associated with immunoglobulin E-mediated allergic diseases such as asthma, allergic rhinitis, atopic dermatitis, urticaria, angioedema, and anaphylaxis. However, the SPT is labor-intensive and time-consuming due to the necessity of measuring the sizes of the erythema and wheals induced by allergens on the skin. In this study, we used an image preprocessing method and a deep learning model to segment wheals and erythema in SPT images captured by a smartphone camera. Subsequently, we assessed the deep learning model's performance by comparing the results with ground-truth data. Using contrast-limited adaptive histogram equalization (CLAHE), an image preprocessing technique designed to enhance image contrast, we augmented the chromatic contrast in 46 SPT images from 33 participants. We established a deep learning model for wheal and erythema segmentation using 144 and 150 training datasets, respectively. The wheal segmentation model achieved an accuracy of 0.9985, a sensitivity of 0.5621, a specificity of 0.9995, and a Dice similarity coefficient of 0.7079, whereas the erythema segmentation model achieved an accuracy of 0.9660, a sensitivity of 0.5787, a specificity of 0.97977, and a Dice similarity coefficient of 0.6636. The use of image preprocessing and deep learning technology in SPT is expected to have a significant positive impact on medical practice by ensuring the accurate segmentation of wheals and erythema, producing consistent evaluation results, and simplifying diagnostic processes.

Intraoperative detection of parathyroid glands using artificial intelligence: optimizing medical image training with data augmentation methods.

BACKGROUND: Postoperative hypoparathyroidism is a major complication of thyroidectomy, occurring when the parathyroid glands are inadvertently damaged during surgery. Although intraoperative images are rarely used to train artificial intelligence (AI) because of its complex nature, AI may be trained to intraoperatively detect parathyroid glands using various augmentation methods. The purpose of this study was to train an effective AI model to detect parathyroid glands during thyroidectomy. METHODS: Video clips of the parathyroid gland were collected during thyroid lobectomy procedures. Confirmed parathyroid images were used to train three types of datasets according to augmentation status: baseline, geometric transformation, and generative adversarial network-based image inpainting. The primary outcome was the average precision of the performance of AI in detecting parathyroid glands. RESULTS: 152 Fine-needle aspiration-confirmed parathyroid gland images were acquired from 150 patients who underwent unilateral lobectomy. The average precision of the AI model in detecting parathyroid glands based on baseline data was 77%. This performance was enhanced by applying both geometric transformation and image inpainting augmentation methods, with the geometric transformation data augmentation dataset showing a higher average precision (79%) than the image inpainting model (78.6%). When this model was subjected to external validation using a completely different thyroidectomy approach, the image inpainting method was more effective (46%) than both the geometric transformation (37%) and baseline (33%) methods. CONCLUSION: This AI model was found to be an effective and generalizable tool in the intraoperative identification of parathyroid glands during thyroidectomy, especially when aided by appropriate augmentation methods. Additional studies comparing model performance and surgeon identification, however, are needed to assess the true clinical relevance of this AI model.

Deep Learning-Based Nystagmus Detection for BPPV Diagnosis.

In this study, we propose a deep learning-based nystagmus detection algorithm using video oculography (VOG) data to diagnose benign paroxysmal positional vertigo (BPPV). Various deep learning architectures were utilized to develop and evaluate nystagmus detection models. Among the four deep learning architectures used in this study, the CNN1D model proposed as a nystagmus detection model demonstrated the best performance, exhibiting a sensitivity of 94.06 ± 0.78%, specificity of 86.39 ± 1.31%, precision of 91.34 ± 0.84%, accuracy of 91.02 ± 0.66%, and an F1-score of 92.68 ± 0.55%. These results indicate the high accuracy and generalizability of the proposed nystagmus diagnosis algorithm. In conclusion, this study validates the practicality of deep learning in diagnosing BPPV and offers avenues for numerous potential applications of deep learning in the medical diagnostic sector. The findings of this research underscore its importance in enhancing diagnostic accuracy and efficiency in healthcare.

Machine learning model based on radiomics features for AO/OTA classification of pelvic fractures on pelvic radiographs.

Depending on the degree of fracture, pelvic fracture can be accompanied by vascular damage, and in severe cases, it may progress to hemorrhagic shock. Pelvic radiography can quickly diagnose pelvic fractures, and the Association for Osteosynthesis Foundation and Orthopedic Trauma Association (AO/OTA) classification system is useful for evaluating pelvic fracture instability. This study aimed to develop a radiomics-based machine-learning algorithm to quickly diagnose fractures on pelvic X-ray and classify their instability. data used were pelvic anteroposterior radiographs of 990 adults over 18 years of age diagnosed with pelvic fractures, and 200 normal subjects. A total of 93 features were extracted based on radiomics:18 first-order, 24 GLCM, 16 GLRLM, 16 GLSZM, 5 NGTDM, and 14 GLDM features. To improve the performance of machine learning, the feature selection methods RFE, SFS, LASSO, and Ridge were used, and the machine learning models used LR, SVM, RF, XGB, MLP, KNN, and LGBM. Performance measurement was evaluated by area under the curve (AUC) by analyzing the receiver operating characteristic curve. The machine learning model was trained based on the selected features using four feature-selection methods. When the RFE feature selection method was used, the average AUC was higher than that of the other methods. Among them, the combination with the machine learning model SVM showed the best performance, with an average AUC of 0.75±0.06. By obtaining a feature-importance graph for the combination of RFE and SVM, it is possible to identify features with high importance. The AO/OTA classification of normal pelvic rings and pelvic fractures on pelvic AP radiographs using a radiomics-based machine learning model showed the highest AUC when using the SVM classification combination. Further research on the radiomic features of each part of the pelvic bone constituting the pelvic ring is needed.

A Method to Reduce Tension Differences in Pull/Push Manipulation for a Robot in Fluorescence Emission-Guided Surgical Microscopy.

MOTIVATION: A fluorescence emission-guided microscope used to monitor the outcome of cancer removal surgery is highly effective when employing a manipulator to motorize and switch the observation direction. It is necessary to minimize the alignment of looper tension between the stands for pull/push to change the direction of the manipulator and reduce the error rate caused by tension differences. This paper presents a method to minimize the error rate of looper tension between the stands. METHODS: \The looper is inserted between the stands of the manipulator to minimize the difference in tension and make the stress on the pull and push of the looper constant. The constant stress allows the manipulator to move stably in left/right, up/down, and left/right movements, which will be effective for full-camera observation and close-up shots of the end effector. RESULTS: Reducing the tolerance for differences in the manipulator's looper tension (angle and tension) is crucial. When the input value of the looper tension angle is 50°, the output should closely match 50°. Consequently, the measured response has a tolerance of ±49.98%, resulting in an error rate of .02% (1/50th level). CONCLUSION: A method is proposed to minimize the error rate of the manipulator's looper tension in a robot-based fluorescence emission-guided microscope used to observe the status of cancer surgery. As a result, a stable manipulator with a minimal error rate can achieve a 3.986x magnification for close-up observation by switching between high and low orientations.

Deep-learning-based pelvic automatic segmentation in pelvic fractures.

With the recent increase in traffic accidents, pelvic fractures are increasing, second only to skull fractures, in terms of mortality and risk of complications. Research is actively being conducted on the treatment of intra-abdominal bleeding, the primary cause of death related to pelvic fractures. Considerable preliminary research has also been performed on segmenting tumors and organs. However, studies on clinically useful algorithms for bone and pelvic segmentation, based on developed models, are limited. In this study, we explored the potential of deep-learning models presented in previous studies to accurately segment pelvic regions in X-ray images. Data were collected from X-ray images of 940 patients aged 18 or older at Gachon University Gil Hospital from January 2015 to December 2022. To segment the pelvis, Attention U-Net, Swin U-Net, and U-Net were trained, thereby comparing and analyzing the results using five-fold cross-validation. The Swin U-Net model displayed relatively high performance compared to Attention U-Net and U-Net models, achieving an average sensitivity, specificity, accuracy, and dice similarity coefficient of 96.77%, of 98.50%, 98.03%, and 96.32%, respectively.

Deep Learning-Based Multi-Class Segmentation of the Paranasal Sinuses of Sinusitis Patients Based on Computed Tomographic Images.

Accurate paranasal sinus segmentation is essential for reducing surgical complications through surgical guidance systems. This study introduces a multiclass Convolutional Neural Network (CNN) segmentation model by comparing four 3D U-Net variations-normal, residual, dense, and residual-dense. Data normalization and training were conducted on a 40-patient test set (20 normal, 20 abnormal) using 5-fold cross-validation. The normal 3D U-Net demonstrated superior performance with an F1 score of 84.29% on the normal test set and 79.32% on the abnormal set, exhibiting higher true positive rates for the sphenoid and maxillary sinus in both sets. Despite effective segmentation in clear sinuses, limitations were observed in mucosal inflammation. Nevertheless, the algorithm's enhanced segmentation of abnormal sinuses suggests potential clinical applications, with ongoing refinements expected for broader utility.

Development of an Artificial Soft Solid Gel Using Gelatin Material for High-Quality Ultrasound Diagnosis.

For ultrasound diagnosis, a gel is applied to the skin. Ultrasound gel serves to block air exposure and match impedance between the skin and the probe, enhancing imaging efficiency. However, if use of the ultrasound gel exceeds a certain period of time, it may dry out and be exposed to air, causing impedance mismatch and reducing imaging resolution. In such cases, the use of a soft, solid gel proves advantageous, as it can be employed for an extended period without succumbing to the drying phenomenon and can be reused after disinfection. Its soft consistency ensures excellent skin adhesion. Our soft solid gel demonstrated approximately 1.2 times better performance than water, silicone, and traditional ultrasound gels. When comparing the dimensions of grayscale, dead zone, vertical, and horizontal regions, the measurements for the traditional ultrasound gel were 93.79 mm, 45.32 mm, 103.13 mm, 83.86 mm, and 83.86 mm, respectively. In contrast, the proposed soft solid gel exhibited dimensions of 105.64 mm, 34.48 mm, 141.1 mm, and 102.8 mm.

Automatic Segmentation and Radiologic Measurement of Distal Radius Fractures Using Deep Learning.

Background: Recently, deep learning techniques have been used in medical imaging studies. We present an algorithm that measures radiologic parameters of distal radius fractures using a deep learning technique and compares the predicted parameters with those measured by an orthopedic hand surgeon. Methods: We collected anteroposterior (AP) and lateral X-ray images of 634 wrists in 624 patients with distal radius fractures treated conservatively with a follow-up of at least 2 months. We allocated 507 AP and 507 lateral images to the training set (80% of the images were used to train the model, and 20% were utilized for validation) and 127 AP and 127 lateral images to the test set. The margins of the radius and ulna were annotated for ground truth, and the scaphoid in the lateral views was annotated in the box configuration to determine the volar side of the images. Radius segmentation was performed using attention U-Net, and the volar/dorsal side was identified using a detection and classification model based on RetinaNet. The proposed algorithm measures the radial inclination, dorsal or volar tilt, and radial height by index axes and points from the segmented radius and ulna. Results: The segmentation model for the radius exhibited an accuracy of 99.98% and a Dice similarity coefficient (DSC) of 98.07% for AP images, and an accuracy of 99.75% and a DSC of 94.84% for lateral images. The segmentation model for the ulna showed an accuracy of 99.84% and a DSC of 96.48%. Based on the comparison of the radial inclinations measured by the algorithm and the manual method, the Pearson correlation coefficient was 0.952, and the intraclass correlation coefficient was 0.975. For dorsal/volar tilt, the correlation coefficient was 0.940, and the intraclass correlation coefficient was 0.968. For radial height, it was 0.768 and 0.868, respectively. Conclusions: The deep learning-based algorithm demonstrated excellent segmentation of the distal radius and ulna in AP and lateral radiographs of the wrist with distal radius fractures and afforded automatic measurements of radiologic parameters.

A Composite Blood Biomarker Including AKR1B10 and Cytokeratin 18 for Progressive Types of Nonalcoholic Fatty Liver Disease.

BACKGRUOUND: We aimed to evaluate whether composite blood biomarkers including aldo-keto reductase family 1 member B10 (AKR1B10) and cytokeratin 18 (CK-18; a nonalcoholic steatohepatitis [NASH] marker) have clinically applicable performance for the diagnosis of NASH, advanced liver fibrosis, and high-risk NASH (NASH+significant fibrosis). METHODS: A total of 116 subjects including healthy control subjects and patients with biopsy-proven nonalcoholic fatty liver disease (NAFLD) were analyzed to assess composite blood-based and imaging-based biomarkers either singly or in combination. RESULTS: A composite blood biomarker comprised of AKR1B10, CK-18, aspartate aminotransferase (AST), and alanine aminotransferase (ALT) showed excellent performance for the diagnosis of, NASH, advanced fibrosis, and high-risk NASH, with area under the receiver operating characteristic curve values of 0.934 (95% confidence interval [CI], 0.888 to 0.981), 0.902 (95% CI, 0.832 to 0.971), and 0.918 (95% CI, 0.862 to 0.974), respectively. However, the performance of this blood composite biomarker was inferior to that various magnetic resonance (MR)-based composite biomarkers, such as proton density fat fraction/MR elastography- liver stiffness measurement (MRE-LSM)/ALT/AST for NASH, MRE-LSM+fibrosis-4 index for advanced fibrosis, and the known MR imaging-AST (MAST) score for high-risk NASH. CONCLUSION: Our blood composite biomarker can be useful to distinguish progressive forms of NAFLD as an initial noninvasive test when MR-based tools are not available.

Predicting Mismatch Repair Deficiency Status in Endometrial Cancer through Multi-Resolution Ensemble Learning in Digital Pathology.

For molecular classification of endometrial carcinoma, testing for mismatch repair (MMR) status is becoming a routine process. Mismatch repair deficiency (MMR-D) is caused by loss of expression in one or more of the 4 major MMR proteins: MLH1, MSH2, MSH6, PHS2. Over 30% of patients with endometrial cancer have MMR-D. Determining the MMR status holds significance as individuals with MMR-D are potential candidates for immunotherapy. Pathological whole slide image (WSI) of endometrial cancer with immunohistochemistry results of MMR proteins were gathered. Color normalization was applied to the tiles using a CycleGAN-based network. The WSI was divided into tiles at three different magnifications (2.5 × , 5 × , and 10 ×). Three distinct networks of the same architecture were employed to include features from all three magnification levels and were stacked for ensemble learning. Three architectures, InceptionResNetV2, EfficientNetB2, and EfficientNetB3 were employed and subjected to comparison. The per-tile results were gathered to classify MMR status in the WSI, and prediction accuracy was evaluated using the following performance metrics: AUC, accuracy, sensitivity, and specificity. The EfficientNetB2 was able to make predictions with an AUC of 0.821, highest among the three architectures, and an overall AUC range of 0.767 - 0.821 was reported across the three architectures. In summary, our study successfully predicted MMR classification from pathological WSIs in endometrial cancer through a multi-resolution ensemble learning approach, which holds the potential to facilitate swift decisions on tailored treatment, such as immunotherapy, in clinical settings.

Automated Segmentation and Diagnostic Measurement for the Evaluation of Cervical Spine Injuries Using X-Rays.

Accurate assessment of cervical spine X-ray images through diagnostic metrics plays a crucial role in determining appropriate treatment strategies for cervical injuries and evaluating surgical outcomes. Such assessment can be facilitated through the use of automatic methods such as machine learning and computer vision algorithms. A total of 852 cervical X-rays obtained from Gachon Medical Center were used for multiclass segmentation of the craniofacial bones (hard palate, basion, opisthion) and cervical spine (C1-C7), incorporating architectures such as EfficientNetB4, DenseNet201, and InceptionResNetV2. Diagnostic metrics automatically measured using computer vision algorithms were compared with manually measured metrics through Pearson's correlation coefficient and paired t-tests. The three models demonstrated high average dice coefficient values for the cervical spine (C1, 0.93; C2, 0.96; C3, 0.96; C4, 0.96; C5, 0.96; C6, 0.96; C7, 0.95) and lower values for the craniofacial bones (hard palate, 0.69; basion, 0.81; opisthion, 0.71). Comparison of manually measured metrics and automatically measured metrics showed high Pearson's correlation coefficients in McGregor's line (r = 0.89), space available cord (r = 0.94), cervical sagittal vertical axis (r = 0.99), cervical lordosis (r = 0.88), lower correlations in basion-dens interval (r = 0.65), basion-axial interval (r = 0.72), and Powers ratio (r = 0.62). No metric showed adjusted significant differences at P < 0.05 between manual and automatic metric measuring methods. These findings demonstrate the potential of multiclass segmentation in automating the measurement of diagnostic metrics for cervical spine injuries and showcase the clinical potential for diagnosing cervical spine injuries and evaluating cervical surgical outcomes.

A Comparative Study of Performance Between Federated Learning and Centralized Learning Using Pathological Image of Endometrial Cancer.

Federated learning, an innovative artificial intelligence training method, offers a secure solution for institutions to collaboratively develop models without sharing raw data. This approach offers immense promise and is particularly advantageous for domains dealing with sensitive information, such as patient data. However, when confronted with a distributed data environment, challenges arise due to data paucity or inherent heterogeneity, potentially impacting the performance of federated learning models. Hence, scrutinizing the efficacy of this method in such intricate settings is indispensable. To address this, we harnessed pathological image datasets of endometrial cancer from four hospitals for training and evaluating the performance of a federated learning model and compared it with a centralized learning model. With optimal processing techniques (data augmentation, color normalization, and adaptive optimizer), federated learning exhibited lower precision but higher recall and Dice similarity coefficient (DSC) than centralized learning. Hence, considering the critical importance of recall in the context of medical image processing, federated learning is demonstrated as a viable and applicable approach in this field, offering advantages in terms of both performance and data security.

Development and validation of artificial intelligence-based analysis software to support screening system of cervical intraepithelial neoplasia.

Cervical cancer, the fourth most common cancer among women worldwide, often proves fatal and stems from precursor lesions caused by high-risk human papillomavirus (HR-HPV) infection. Accurate and early diagnosis is crucial for effective treatment. Current screening methods, such as the Pap test, liquid-based cytology (LBC), visual inspection with acetic acid (VIA), and HPV DNA testing, have limitations, requiring confirmation through colposcopy. This study introduces CerviCARE AI, an artificial intelligence (AI) analysis software, to address colposcopy challenges. It automatically analyzes Tele-cervicography images, distinguishing between low-grade and high-grade lesions. In a multicenter retrospective study, CerviCARE AI achieved a remarkable sensitivity of 98% for high-risk groups (P2, P3, HSIL or higher, CIN2 or higher) and a specificity of 95.5%. These findings underscore CerviCARE AI's potential as a valuable diagnostic tool for highly accurate identification of cervical precancerous lesions. While further prospective research is needed to validate its clinical utility, this AI system holds promise for improving cervical cancer screening and lessening the burden of this deadly disease.

Comparative Sensing and Judgment Control System for Temperature Maintenance for Optimal Treatment in Hyperthermic Intraperitoneal Chemotherapy Surgery.

For tumors wherein cancer cells remain in the tissue after colorectal cancer surgery, a hyperthermic anticancer agent is injected into the abdominal cavity to necrotize the remaining cancer cells with heat using a hyperthermic intraperitoneal chemotherapy system. However, during circulation, the processing temperature is out of range and the processing result is deteriorated. This paper proposes a look-up table (LUT) module design method that can stably maintain the processing temperature range during circulation via feedback. If the temperature decreases or increases, the LUT transmits a command signal to the heat exchanger to reduce or increase heat input, thereby maintaining the treatment temperature range. The command signal for increasing and decreasing heat input is Tp and Ta, respectively. The command signal for the treatment temperature range is Ts. If drug temperatures below 41 and above 43 °C are input to the LUT, it sends a Tp or Ta signal to the heat exchanger to increase or decrease the input heat, respectively. If the drug's temperature is 41-43 °C, the LUT generates a Ts signal and proceeds with the treatment. The proposed system can automatically control drug temperature using temperature feedback to ensure rapid, accurate, and safe treatment.

Adjustment of the Length Variation With Wire-driven Using the Stand Looper Tension Technique for Surgical Robot Applications.

MOTIVATION: The wire-driven method used in the field of surgical robots has the advantage of light weight. However, in the process of pull and push for the operation of forceps, the length of the wire is not match, causing malfunction. To solve this problem, the application of looper-tension technology would be suitable. This paper contributes to adjusting the length of the wire by inserting a stand between the wire-driven joints and adding a looper-tension between the stands to adjust the rotation radius of the roll. METHODS: The method consisting of three rolls and loopers for connection between the stands minimizes errors by adjusting the length of the loop in a balanced state due to the rotation change of the roll during the pull and push of the robot arm. The angle and tension applied to the looper are 25° and 8.6 MPa, respectively. RESULTS: An output response can be obtained when the reference operating point fluctuates by ± 50% of the input angle and tension, and if the reference operating point fluctuates by ± 30% while the input angle and tension are fixed, the output response occurs oppositely. When a .15 kg object is loaded up/down with 1.5 newton using forceps, the change in length of pull and push coincides. CONCLUSION: The advantage is that the error of wire pull, and push operation can be reduced, and accurate operation can be expected. Since the proposed technology is applied between joints, the integration process is not complicated, and the weight is light.

Machine learning based prediction of recurrence after curative resection for rectal cancer.

PURPOSE: Patients with rectal cancer without distant metastases are typically treated with radical surgery. Post curative resection, several factors can affect tumor recurrence. This study aimed to analyze factors related to rectal cancer recurrence after curative resection using different machine learning techniques. METHODS: Consecutive patients who underwent curative surgery for rectal cancer between 2004 and 2018 at Gil Medical Center were included. Patients with stage IV disease, colon cancer, anal cancer, other recurrent cancer, emergency surgery, or hereditary malignancies were excluded from the study. The Synthetic Minority Oversampling Technique with Tomek link (SMOTETomek) technique was used to compensate for data imbalance between recurrent and no-recurrent groups. Four machine learning methods, logistic regression (LR), support vector machine (SVM), random forest (RF), and Extreme gradient boosting (XGBoost), were used to identify significant factors. To overfit and improve the model performance, feature importance was calculated using the permutation importance technique. RESULTS: A total of 3320 patients were included in the study. After exclusion, the total sample size of the study was 961 patients. The median follow-up period was 60.8 months (range:1.2-192.4). The recurrence rate during follow-up was 13.2% (n = 127). After applying the SMOTETomek method, the number of patients in both groups, recurrent and non-recurrent group were equalized to 667 patients. After analyzing for 16 variables, the top eight ranked variables {pathologic Tumor stage (pT), sex, concurrent chemoradiotherapy, pathologic Node stage (pN), age, postoperative chemotherapy, pathologic Tumor-Node-Metastasis stage (pTNM), and perineural invasion} were selected based on the order of permutational importance. The highest area under the curve (AUC) was for the SVM method (0.831). The sensitivity, specificity, and accuracy were found to be 0.692, 0.814, and 0.798, respectively. The lowest AUC was obtained for the XGBoost method (0.804), with a sensitivity, specificity, and accuracy of 0.308, 0.928, and 0.845, respectively. The variable with highest importance was pT as assessed through SVM, RF, and XGBoost (0.06, 0.12, and 0.13, respectively), whereas pTNM had the highest importance when assessed by LR (0.05). CONCLUSIONS: In the current study, SVM showed the best AUC, and the most influential factor across all machine learning methods except LR was found to be pT. The rectal cancer patients who have a high pT stage during postoperative follow-up are need to be more close surveillance.