The present study aimed to develop a machine learning network to diagnose middle ear diseases with tympanic membrane images and to identify its assistive role in the diagnostic process. The medical records of subjects who underwent ear endoscopy tests were reviewed. From these records, 2272 diagnostic tympanic membranes images were appropriately labeled as normal, otitis media with effusion (OME), chronic otitis media (COM), or cholesteatoma and were used for training. We developed the “ResNet18 + Shuffle” network and validated the model performance. Seventy-one representative cases were selected to test the final accuracy of the network and resident physicians. We asked 10 resident physicians to make diagnoses from tympanic membrane images with and without the help of the machine learning network, and the change of the diagnostic performance of resident physicians with the aid of the answers from the machine learning network was assessed. The devised network showed a highest accuracy of 97.18%. A five-fold validation showed that the network successfully diagnosed ear diseases with an accuracy greater than 93%. All resident physicians were able to diagnose middle ear diseases more accurately with the help of the machine learning network. The increase in diagnostic accuracy was up to 18% (1.4% to 18.4%). The machine learning network successfully classified middle ear diseases and was assistive to clinicians in the interpretation of tympanic membrane images.
1 January 2004 in Journal of Korean Medical Science
Objective: The objective of this prospective study was to assess longitudinal variations in health-related quality of life (HR-QOL) in patients diagnosed with differentiated thyroid cancer (DTC) before and after thyroidectomy. Methods: A cohort of 185 DTC patients who underwent thyroidectomy between January 2013 and December 2017 and who completed all necessary questionnaires was evaluated. Their HR-QOL was gauged using the University of Washington Quality of Life questionnaire (UW-QOL) and the City of Hope Quality of Life-Thyroid Version questionnaire (QOL-TV) both prior to surgery and at 3 months, 6 months, 1 year, 2 years, 3 years, and 5 years postoperatively. Results: Out of 185 patients, 150 (81.1%) were female, with an average age of 48.7 ± 12.9 years. For both UW-QOL and QOL-TV, the total composite QOL scores notably declined from preoperative levels to 3 months postoperatively, then gradually improved over 5 years, ultimately exceeding preoperative scores. Factors such as total thyroidectomy, radioactive iodine (RAI) ablation, and postoperative hypoparathyroidism were associated with lower physical composite QOL scores. Patients who underwent remote-access thyroidectomy expressed significantly higher satisfaction with appearance compared with those who had conventional thyroidectomy. Mood and anxiety were major clinical concerns both before and after surgery, showing considerable improvement postoperatively. Conclusion: For DTC patients, HR-QOL experienced a significant drop 3 months postsurgery, subsequently showing gradual improvement, surpassing preoperative QOL by 5 years. Factors contributing to improved physical QOL included the utilization of remote-access thyroidectomy, less extensive thyroidectomy, and the absence of RAI ablation and hypoparathyroidism.
1 November 2020 in Clinical and Experimental Otorhinolaryngology
Objectives. This study was conducted evaluate the efficacy of electromyography (EMG) using transcartilaginous (TC) electrodes through the thyroid cartilage and perichondrium.Methods. We prospectively collected EMG data from intraoperative neuromonitoring (IONM) of 54 nerves at risk in 38 patients during thyroidectomy. We followed standardized IONM procedures in all operations. EMG signals from both endotracheal tube (ET) electrodes and TC needle electrodes were recorded simultaneously. We compared the characteristics of the EMG signals and the efficacy of both methods.Results. Significantly higher mean EMG amplitudes were recorded by TC electrodes than by ET electrodes in all four-step procedures (V1-R1-R2-V2, P<0.001). Loss of signal (LOS) occurred in five patients in ET electrodes, but in only two patients in TC electrodes. Postoperative laryngoscopy revealed recurrent laryngeal nerve palsy in the two patients who showed LOS from both the ET and TC electrodes, and vocal cord movement was intact in the other three patients. Therefore, the positive predictive values of LOS in ET and TC electrodes were 40% and 100%, respectively.Conclusion. EMG recording using TC needle electrodes is feasible and effective, making it a good alternative technique for IONM.