Research

The experience of pain is quite variable across individuals. How do an individual’s cognitive and affective states evaluate and modulate nociceptive information?

​

In a delayed discrimination paradigm, we discovered that participants’ ability to encode the intensity information of pain was biased by their anxiety and involved a distinct neural stream encompassing the medial thalamus, medial prefrontal cortex, and amygdala (Tseng et al., 2017, The Journal of Neuroscience). When participants discriminated painful stimulus intensities, their discrimination ability was modulated by their vigilance to pain. This modulation was mirrored by an enhanced coupling within prefrontal-thalamic attention networks, which allow the superior prefrontal region to estimate the relative intensity differences between noxious stimuli (Yang et al.,2018, Pain). These studies exemplify how human responses to pain are modulated by cognition and emotion.

Our subjective experience of pain is largely shaped by expectations. How do top-down expectations interact with bottom-up nociceptive inputs to modulate pain perception?

In a stimulus expectancy paradigm, we found that positive expectations (expectations of decreased pain) and negative expectations (expectations of increased pain) engaged separate brain regions encoding the mismatch between actual and expected pain (i.e., prediction error) and involved opposite coupling with the descending pain modulatory system. Interestingly, positive and negative expectations produced correlated pain rating changes and brain activation. These findings suggest that aversive prediction error signals underlie stimulus expectancy effects on pain, and positive expectation- and negative expectation-related modulation mechanisms are interrelated. These mechanisms help to explain why we humans can adapt quickly and appropriately to noxious stimuli whose intensity deviates from our expectations. (Shih et al., 2019, The Journal of Neuroscience).

In another stimulus expectancy paradigm combining emotion regulation and computational modeling, we revealed that the relative emotional responses between expectations and sensory inputs modulated their integration to form pain perception in healthy subjects. These phenomena involved brain regions processing anxiety for negative expectations and processing pleasantness for positive expectations. For negative expectations, subjects’ anxiety also appeared to impair the updating of expectations via suppressed prediction error signals, thus perpetuating negative expectancy effects. This study highlights the important role of emotions when pain experiences are shaped by stimulus expectancy, and adds to the growing studies investigating the computational mechanisms behind pain modulation by stimulus expectancy in humans. (Tsai et al., 2024, The Journal of Neuroscience).

While perceiving somatic stimuli, the ability to accurately discriminate their spatial and temporal properties is essential for human behavior, but how does our brain differentiate these properties?

In a tactile discrimination paradigm, we found that tactile detection elicited activation specifically involved in spatial discrimination within the right inferior and superior parietal lobules, and the functional connectivity between these two regions predicted individual spatial discriminability. In stark contrast with spatial discrimination, tactile detection produced little activation specifically related to temporal discrimination. The revealed close relationship between detection and spatial discrimination not only elucidates the substantial influence of sensory detection on spatial information processing, but aids in explaining why we can promptly and precisely direct our response to a somatic stimulus from the external world (Huang et al., 2022, Cerebral Cortex).

We applied this fMRI paradigm to investigate the mechanism underpinning somatosensory aberrations in patients with Parkinson’s disease (PD). Our work showed that reduced functional connectivity between posterior putamen (pPut) and inferior parietal lobule (IPL) predicted decreased IPL activity during tactile detection, which hypoactivation negatively correlated with tactile detection threshold in PD patients. Importantly, gait training with rhythmic auditory stimulation for 3 weeks normalized the pPut-IPL dysconnectivity and IPL hypoactivity as well as tactile detection abnormalities. Training-induced increases in the strength of pPut-IPL coupling predicted the change in IPL activation during tactile detection. These findings indicate that pathological spread from striatal to parietal cortical regions may underlie reduced tactile detectability in PD. Rhythmic auditory-motor intervention acts as an effective strategy to improve patients’ deficits in somatosensation through restoring striatal-cortical dysconnectivity (Huang et al., 2025, Neurobiology of Disease).

Habituation to painful stimuli is an endogenous pain alleviation mechanism, and reduced pain habituation has been observed in many chronic pain conditions. In ethology, animals show decreased fear responses when habituating to repeated threats. Does a similar mechanism operate in human pain habituation?

Our lab explored these mechanisms from a motivational-ethological perspective. In a recent study, we found that pain habituation at a repetitively stimulated forearm site involved reduced fear and engaged brain regions linked to fear reduction, including the amygdala, anterior cingulate, and ventromedial prefrontal cortex (vmPFC). Individual pain-related fear predicted activity in the periaqueductal gray, a pain-modulating center, which covaried with vmPFC responsivity. Habituation also occurred at non-stimulated sites, with its extent predicted by habituation at the stimulated site, a process involving the vmPFC. These findings suggest that fear reduction contributes to pain habituation, shedding light on why impaired fear reduction and diminished pain habituation often coexist in chronic pain conditions (Lin et al., 2025, Psychophysiology).

How does the brain integrate reward and punishment experiences when the same choice can lead to different outcomes?

We study how the brain integrates reward and punishment experiences to guide learning and decision-making. In everyday life, the same choice can lead to different outcomes depending on the situation. For example, riding a bicycle to work may save time and be rewarding on a sunny day, but the same choice may result in an unpleasant experience when it rains. Although the bicycle remains a good option when the weather is favorable, the memory of getting soaked may make us hesitate, avoid the choice, or consider other options the next time we see dark clouds. Our research shows that past punishment associated with a choice can increase exploration of alternative options and interfere with subsequent reward learning, an effect modulated by dopamine. These findings reveal how the brain integrates competing value signals to guide behavior and may help explain learning and decision-making disturbances in dopamine-related neuropsychiatric conditions. (Lin et al., 2026, PLOS Biology)(NTU News Letter).