Development of a spontaneous pain indicator based on brain cellular calcium using deep learning

Abstract

Chronic pain remains an intractable condition in millions of patients worldwide. Spontaneous ongoing pain is a major clinical problem of chronic pain and is extremely challenging to diagnose and treat compared to stimulus-evoked pain. Although extensive efforts have been made in preclinical studies, there still exists a mismatch in pain type between the animal model and humans (i.e., evoked vs. spontaneous), which obstructs the translation of knowledge from preclinical animal models into objective diagnosis and effective new treatments. Here, we developed a deep learning algorithm, designated AI-bRNN (Average training, Individual test-bidirectional Recurrent Neural Network), to detect spontaneous pain information from brain cellular Ca2+ activity recorded by two-photon microscopy imaging in awake, head-fixed mice. AI-bRNN robustly determines the intensity and time points of spontaneous pain even in chronic pain models and evaluates the efficacy of analgesics in real time. Furthermore, AI-bRNN can be applied to various cell types (neurons and glia), brain areas (cerebral cortex and cerebellum) and forms of somatosensory input (itch and pain), proving its versatile performance. These results suggest that our approach offers a clinically relevant, quantitative, real-time preclinical evaluation platform for pain medicine, thereby accelerating the development of new methods for diagnosing and treating human patients with chronic pain.

Machine learning-based evaluation of spontaneous pain and analgesics from cellular calcium signals in the mouse primary somatosensory cortex using explainable features

“…In previous research, our team developed an algorithm that uses DL to detect pain from S1 calcium signals (Yoon et al, 2022 ). In the current study, we introduced an ML model that, while aiming to achieve similar goals as our prior research, offers a range of distinct benefits.…”

“…To establish the complete Freund's adjuvant (CFA) model (Yoon et al, 2022 ), 10 μl of CFA were injected subcutaneously into the plantar surface of the right hind paw. To establish the chemotherapy-induced peripheral neuropathy model (Kim W. et al, 2016 ), oxaliplatin (6 mg/kg) was treated intraperitoneally.…”

“…Leveraging advancements in computational techniques, ML has become a key tool in pain research, outperforming prior methods in analyzing complex calcium imaging data (Singh et al, 2016 ; Boissoneault et al, 2017 ; Lötsch and Ultsch, 2018 ; Jones et al, 2020 ). Previously, we developed a DL model for assessing spontaneous pain (Yoon et al, 2022 ). DL is adept at detecting complex patterns in pain data, often missed by human analysis, thereby offering a more explicit detection of pain.…”

“…Our primary goal in this study is to offer a more transparent and interpretable method for measuring spontaneous pain in preclinical research. Unlike previous research that predominantly relied on DL with black-box mechanisms (Yoon et al, 2022 ), we adopt a ML approach that emphasizes manual feature extraction. This methodology aims to provide a clearer understanding of the reasons behind the classification of S1 calcium signals as “pain” or “no pain.” Furthermore, by extending beyond the traditional binary classification and introducing a third-class label, we seek to achieve a more detailed interpretation of neural activity patterns, discerning whether the neural activity pattern induced by candidate drugs mirrors existing analgesics or introduces a new signature.…”

Multiplexed Representation of Itch and Pain and Their Interaction in the Primary Somatosensory Cortex

“…These results suggest that S1 can antagonistically modulate itch and pain [ 20 , 21 ]. In our previous study, we showed that itch and pain could be distinguished by a deep-learning algorithm based on S1 neuronal activity, implying that S1 neurons have a distinct response pattern for itch and pain [ 22 ]. However, how S1 neurons differentially encode these two distinct sensations and whether S1 represents the interaction between two sensations remains unexplored at the cellular level.…”