OBJECTIVE: Hand function is central to inter- actions with our environment. Developing a comprehen- sive model of hand grasps in naturalistic environments is crucial across various disciplines, including robotics, ergonomics, and rehabilitation. Creating such a taxonomy poses challenges due to the significant variation in grasping strategies that individuals may employ. For instance, individuals with impaired hands, such as those with spinal cord injuries (SCI), may develop unique grasps not used by unimpaired individuals. These grasping techniques may differ from person to person, influenced by variable senso- rimotor impairment, creating a need for personalized meth- ods of analysis. METHOD: This study aimed to automatically identify the dominant distinct hand grasps for each indi- vidual without reliance on a priori taxonomies, by applying semantic clustering to egocentric video. Egocentric video recordings collected in the homes of 19 individual with cervical SCI were used to cluster grasping actions with semantic significance. A deep learning model integrating posture and appearance data was employed to create a personalized hand taxonomy. RESULTS: Quantitative analysis reveals a cluster purity of 67.6% ± 24.2% with 18.0% ± 21.8% redundancy. Qualitative assessment revealed meaningful clusters in video content. DISCUSSION: This methodology provides a flexible and effective strategy to analyze hand function in the wild, with applications in clinical assess- ment and in-depth characterization of human-environment interactions in a variety of contexts.

OBJECTIVE: The hand function of individuals with spinal cord injury (SCI) plays a crucial role in their independence and quality of life. Wearable cameras provide an opportunity to analyze hand function in non-clinical environments. Summarizing the video data and documenting dominant hand grasps and their usage frequency would allow clinicians to quickly and precisely analyze hand function. METHOD: We introduce a new hierarchical model to summarize the grasping strategies of individuals with SCI at home. The first level classifies hand-object interaction using hand-object contact estimation. We developed a new deep model in the second level by incorporating hand postures and hand-object contact points using contextual information. RESULTS: In the first hierarchical level, a mean of 86% ±1.0% was achieved among 17 participants. At the grasp classification level, the mean average accuracy was 66.2 ±12.9%. The grasp classifier's performance was highly dependent on the participants, with accuracy varying from 41% to 78%. The highest grasp classification accuracy was obtained for the model with smoothed grasp classification, using a ResNet50 backbone architecture for the contextual head and a temporal pose head. DISCUSSION: We introduce a novel algorithm that, for the first time, enables clinicians to analyze the quantity and type of hand movements in individuals with spinal cord injury at home. The algorithm can find applications in other research fields, including robotics, and most neurological diseases that affect hand function, notably, stroke and Parkinson's.

Egocentric vision (a.k.a. first-person vision-FPV) applications have thrived over the past few years, thanks to the availability of affordable wearable cameras and large annotated datasets. The position of the wearable camera (usually mounted on the head) allows recording exactly what the camera wearers have in front of them, in particular hands and manipulated objects. This intrinsic advantage enables the study of the hands from multiple perspectives: localizing hands and their parts within the images; understanding what actions and activities the hands are involved in; and developing human-computer interfaces that rely on hand gestures. In this survey, we review the literature that focuses on the hands using egocentric vision, categorizing the existing approaches into: localization (where are the hands or parts of them?); interpretation (what are the hands doing?); and application (e.g., systems that used egocentric hand cues for solving a specific problem). Moreover, a list of the most prominent datasets with hand-based annotations is provided.

OBJECTIVE: Cervical spinal cord injury (cSCI) can impair motor function in the upper limbs. Video from wearable cameras (egocentric video) has the potential to provide monitoring of rehabilitation outcomes at home, but methods for automated analysis of this data are needed. Wrist flexion and extension are essential elements to track grasping strategies after cSCI, as they may reflect the use of the tenodesis grasp, a common compensatory strategy. However, there is no established method to evaluate wrist flexion and extension from egocentric video. METHODS: We propose a machine-learning-based approach comprising three steps-hand detection, pose estimation, and arm orientation estimation-to estimate wrist angle data, leading to the detection of tenodesis grasp. RESULTS: The hand detection in conjunction with the pose estimation algorithm correctly located wrist and index finger metacarpophalangeal coordinates in 63% and 76% of 15,319 annotated frames, respectively, extracted from egocentric videos of individuals with cSCI performing activities of daily living in a home simulation laboratory. The arm orientation algorithm had a mean absolute error of 2.76 +/- 0.39 degrees in 12,863 labeled frames. Using these estimates, the presence of a tenodesis grasp was correctly detected in 72% +/- 11% of frames in videos of 6 activities. CONCLUSION: The results provided a clear indication of which participants relied on tenodesis grasp and which did not. SIGNIFICANCE: This paradigm provides the first method that can enable clinicians and researchers to monitor the use of the tenodesis grasp by individuals with cSCI at home, with implications for remote therapeutic guidance.

In order to develop effective interventions for restoring upper extremity function after cervical spinal cord injury, tools are needed to accurately measure hand function throughout the rehabilitation process. However, there is currently no suitable method to collect information about hand function in the community, when patients are not under direct observation of a clinician. We propose a wearable system that can monitor functional hand use using computer vision techniques applied to egocentric camera videos. To this end, in this study we demonstrate the feasibility of detecting interactions of the hand with objects in the environment from egocentric video. The system consists of a preprocessing step where the hand is segmented out from the background. The algorithm then extracts features associated with hand-object interactions. This includes comparing motion cues in the region near the hand (i.e., where the object is most likely to be located) to the motion of the hand itself, as well as to the motion of the background. Features representing hand shape are also extracted. The features serve as inputs to a random forest classifier, which was tested with a dataset of 14 activities of daily living as well as noninteractive tasks in five environment (total video duration of 44.16 min). The average F-score for the classifier was 0.85 for leave-one-activity out in our dataset set and 0.91 for a publicly available set (1.72 min) when filtered with a moving average. These results suggest that using egocentric video to monitor functional hand use at home is feasible.