Evelyn Maniaki: Accelerometer-derived classifiers for early detection of degenerative joint disease in cats

(McIlwraith and VACHON, 1988; Lascelles, 2010). This deterioration goes beyond the natural wear and tear associated with aging, leading to pain, impaired joint function, and poor quality of life. The condition results in articular cartilage degeneration and pathological changes in periarticular tissues, which cause low grade inflammation in the joint and reduced range of motion together with pain on movement. Known risk factors for DJD in cats include late neutering, obesity, outdoor access and trauma (Maniaki et al., 2021; Maniaki, 2020). It is estimated that most older cats have DJD, with 61 % of cats aged 6 years or older showing radiographic evidence of osteoarthritis (Slingerland et al., 2011; Lascelles, 2010).

Decreased mobility is a highly prevalent clinical sign of DJD in cats (Slingerland et al., 2011). Although radiographic evaluation provides an accurate diagnosis and assessment of DJD, identifying pain responses in cats through clinical examination can often be challenging as it heavily relies on interpreting behavioural cues that may not consistently reflect underlying pain.

The subjective nature of pain perception in animals complicates the assessment, as cats may not exhibit overt pain behaviours consistently across different individuals or stages of disease progression. Furthermore, owner separation coupled with the physical examination location can result in clinically significant increases in perceived stress in cats and compromise vital sign assessments (Tateo et al., 2021). Whenever possible, physical examinations and procedures should take place with the owner present and should be conducted with separation from unfamiliar cats to minimize stress and improve the accuracy of clinical evaluations (Griffin et al., 2021).

Moreover, there is no universally accepted, objective method for gauging the pain referable to DJD in cats.

Previous studies have utilised ‘gold standard’ assessment tools, including force plates and pressure mats, which have shown efficacy in detecting mobility impairments (Schnabl-Feichter et al. 2020; Schnabl and Bockstahler, 2015; Stadig and Bergh, 2015). However, these methods are often impractical for routine clinical use in cats due to their unique movement patterns and handling sensitivities.

Although more experienced assessors can accurately perceive pain in cats, the challenge intensifies as radiographic findings do not always consistently align with pain perception, making DJD less obvious to many owners.

Several studies have investigated the use of accelerometers to assess feline mobility and detect changes in activity patterns associated with osteoarthritis. Yamazaki et al. (2020) validated an accelerometer-based activity monitor specifically for cats, demonstrating its accuracy in measuring movement, jumps, and rest periods. Additionally, Yuting (2024) applied machine learning techniques to accelerometer data, successfully identifying behavioural changes in older cats receiving joint supplements. Lascelles et al. (2008) found that activity monitors could detect increased movement following analgesic treatment, suggesting their utility in evaluating pain management strategies. Furthermore, Lascelles (2007) reported a strong correlation between accelerometer-derived activity data and actual distance moved, reinforcing the reliability of these devices for objective mobility assessment.

Presently, there are no fully validated subjective or objective assessment systems for the assessment of chronic DJD-associated pain in cats. Hence, identifying subtle changes in activity patterns represents a potential objective approach for detecting or measuring pain linked to DJD (Slingerland et al., 2011, Lascelles et al., 2010, Lascelles et al., 2012).

The development of an accurate evaluation model could allow earlier detection of DJD in cats, facilitating earlier treatment and potentially improving quality of life. This study aimed to predict early signs of DJD in indoor cats with the use of accelerometers and machine learning techniques. The study hypothesis was that the effect of DJD would be more pronounced during periods of intense activity, such as when the cat performs behaviours such as jumping or moving at high speed as these actions place significant stress on the affected joints. Findings suggest such behaviours, are particularly affected by DJD, Hardie et al. (2002) reported that 90 % of geriatric cats showed radiographic evidence of DJD, with owners observing a reluctance to jump as a primary clinical sign. In Lascelles et al. (2008) study assessing DJD in cats, owners reported that running was one of the activities most affected by the condition.

Read full article here