Comparison among Four Deep Learning Image Classification Algorithms in AI-based Diatom Test.

OBJECTIVES: To select four algorithms with relatively balanced complexity and accuracy among deep learning image classification algorithms for automatic diatom recognition, and to explore the most suitable classification algorithm for diatom recognition to provide data reference for automatic diatom testing research in forensic medicine. METHODS: The "diatom" and "background" small sample size data set (20 000 images) of digestive fluid smear of corpse lung tissue in water were built to train, validate and test four convolutional neural network (CNN) models, including VGG16, ResNet50, InceptionV3 and Inception-ResNet-V2. The receiver operating characteristic curve (ROC) of subjects and confusion matrixes were drawn, recall rate, precision rate, specificity, accuracy rate and F1 score were calculated, and the performance of each model was systematically evaluated. RESULTS: The InceptionV3 model achieved much better results than the other three models with a balanced recall rate of 89.80%, a precision rate of 92.58%. The VGG16 and Inception-ResNet-V2 had similar diatom recognition performance. Although the performance of diatom recall and precision detection could not be balanced, the recognition ability was acceptable. ResNet50 had the lowest diatom recognition performance, with a recall rate of 55.35%. In terms of feature extraction, the four models all extracted the features of diatom and background and mainly focused on diatom region as the main identification basis. CONCLUSIONS: Including the Inception-dependent model, which has stronger directivity and targeting in feature extraction of diatom. The InceptionV3 achieved the best performance on diatom identification and feature extraction compared to the other three models. The InceptionV3 is more suitable for daily forensic diatom examination.

Magnetic resonance imaging: a valuable method for diagnosing chronic lumbago caused by lumbar muscle strain and monitoring healing process / 中华医学杂志(英文版)

<p><b>BACKGROUND</b>Magnetic resonance imaging (MRI) is the most sensitive technique for evaluating the healing process and should be performed before the patients return to their exercise routines. The aim of this research was to diagnose chronic lumbago associated with lumbar muscle strain and to monitor healing process by MRI.</p><p><b>METHODS</b>Sixty-fve symptomatic cases of chronic lumbago caused by lumbar muscle strain were collected from March 2009 to October 2011. MRI was used to examine, diagnose and monitor the healing process. The control group included 65 random cases of asymptomatic volunteers. MRI methods included routine sequences of GRE T1WI, TSE T2WI and special sequences of T2-STIR-FS, combined with DWI. We compared the MRI characteristics of symptomatic cases before and after healing and with asymptomatic controls.</p><p><b>RESULTS</b>The important MRI characteristics of chronic lumbago with lumbar muscle strain included: (1) The low back muscle showed edema. (2) The low back intermuscular spaces showed edema and/or fluid. (3) The low back spaces beside the spinous process showed edema and/or fluid. (4) The low back vertebral articular process fossae or transverse process fossae showed fluid. Of these image characteristics, the intermuscular space edema provided the best diagnostic sensitivity, Se = 83%, with YI = 0.63, p = 74%. The low back muscle edema provided the best diagnostic specificity, Sp = 100%, with YI = 0.66, Π = 83%. And the spaces edema beside the spinous process provided the best diangnostic accuracy, Π = 86%, with YI = 0.71, Se = 80%, Sp = 91%. The diagnosis accurate could be improved by combining multiple MRI characteristics. The diagnostic accuracy could achieve Π = 93%, with YI = 0.86, Se = 100% and Sp = 86% when two characteristics were combined. After rehabilitation care, the edema disappeared on the repeated MRI.</p><p><b>CONCLUSIONS</b>MRI may well be a useful diagnostic method for lumbago with lumbar muscle strain. Combining routine sequences with T2-STIR-FS and DWI sequences could demonstrate the pathological changes of lumbar muscle strain and monitor the healing.</p>