Preprocessing plays a key role in Raman spectral analysis. However, classical preprocessing algorithms often have issues with reducing Raman peak intensities and changing the peak shape when processing spectra. This paper introduces a unified solution for preprocessing based on a convolutional autoencoder to enhance Raman spectroscopy data. One is a denoising algorithm that uses a convolutional denoising autoencoder (CDAE model), and the other is a baseline correction algorithm based on a convolutional autoencoder (CAE+ model). The CDAE model incorporates two additional convolutional layers in its bottleneck layer for enhanced noise reduction. The CAE+ model not only adds convolutional layers at the bottleneck but also includes a comparison function after the decoding for effective baseline correction. The proposed models were validated using both simulated spectra and experimental spectra measured with a Raman spectrometer system. Comparing their performance with that of traditional signal processing techniques, the results of the CDAE-CAE+ model show improvements in noise reduction and Raman peak preservation.
Pathology training resources remain scarce in many parts of the world. With rapid economic development comes the need to educate new pathologists to meet the medical care demands. Our aim was to set up a cost-effective system for training and testing the diagnostic skills of pathologists. Pathologists in nine countries in Asia and South America were invited by the International Society of Urological Pathology (ISUP) to participate in a prostate pathology education course combining image-based tests with lectures and on-line tutorials. The tests and tutorials are available free of charge at the ISUP education website www.edu.isupweb.org. A total of 603 pathologists registered on the website. Of these, 224 completed pre- and post-lecture assessments (tests 1 and 2). Replies were classified as correct/acceptable, when a lesion was accurately classified into clinically relevant categories (benign, cancer, high-grade prostatic intraepithelial neoplasia, intraductal carcinoma of the prostate). The rate of correct/acceptable replies increased from 60.7 to 72.3% in Tests 1 and 2, respectively. In Test 1, pathologists from upper middle, lower middle, and low resource countries gave a correct/acceptable diagnosis in 65.8%, 61.0%, and 47.4%, respectively. Their results improved in Test 2 to 76.4%, 72.5%, and 62.8%, respectively. The greatest improvement in diagnostic ability was achieved in pathologists from the low resource group of countries. The use of web-based testing and training, combined with lectures, is an efficient method for improving diagnostic skills of pathologists in low to middle resource countries.