Search

Pain is an alert state of the human body that can be conveyed to the external world through different modalities. A possible communication channel for human pain is represented by facial expressions, whose role in social interactions has been well established. In this work, the link between pain and transfer entropy (TE), passing through facial expressions, is investigated. A new approach to the vision-based measurement (VBM) of pain is presented, which is based on TE among the time-series of facial landmarks positions. The system is composed of three main blocks: A VBM block for the automatic landmarking and the generation of the time-series from the video-sequences; a second block for the evaluation of TE; and finally a classification model based on machine learning algorithms for pain assessment. A public database of video sequences of patients experiencing pain in a controlled scenario was used for the characterization of the system in terms of accuracy and precision. Different uncertainty contributions related to realistic signals interruptions and fluctuations were modeled and propagated to provide a comprehensive evaluation of the proposed measurement system. The obtained results indicate that TE-based approaches can provide great benefits in automatic pain assessment, opening new perspectives for remote management of patients.

Organs On a Chip (OOCs) represent a sophisticated approach for exploring biological mechanisms and developing therapeutic agents. In conjunction with high-quality time-lapse microscopy (TLM), OOCs allow for the visualization of reconstituted complex biological processes, such as multi-cell-type migration and cell–cell interactions. In this context, increasing the frame rate is desirable to reconstruct accurately cell-interaction dynamics. However, a trade-off between high resolution and carried information content is required to reduce the overall data volume. Moreover, high frame rates increase photobleaching and phototoxicity. As a possible solution for these problems, we report a new hybrid-imaging paradigm based on the integration of OOC/TLMs with a Multi-scale Generative Adversarial Network (GAN) predicting interleaved video frames with the aim to provide high-throughput videos. We tested the performance of the predictive capability of GAN on synthetic videos, as well as on real OOC experiments dealing with tumor–immune cell interactions. The proposed approach offers the possibility to acquire a reduced number of high-quality TLM images without any major loss of information on the phenomena under investigation.

Pain is a health problem of vast dimension that requires prompt diagnosis. However, the evaluation criteria currently used in clinical settings are based on self-rating scales, which are subjective and prone to error. This may cause administering of improper analgesic treatments, either in dosage or in duration. For the scope of providing objective measures of pain and help devising effective therapeutic choices, we designed a personalized platform for sensing and monitoring pain. As input to the platform, visual, speech and physiological clues expressed by the patient during the pain experience are non-invasively measured, then sent to a cloud server to be integrated and analyzed. In this work, preliminary results obtained with the vision-based block of the platform are presented. Because the pain perceived by each individual is highly subjective, once the data are collected from each subject, quantitative descriptors are extracted and personalized selection of relevant features is performed to train a subject-specific regression model. Video sequences of genuine pain felt were tested and weighted accuracy values of 0.81 (0.19) and 0.84 (0.10) were obtained, respectively, for three-class and two-class scenarios, showing benefits and drawbacks of the proposed personalized platform.