We are excited to introduce our latest advancements in preclinical research capabilities powered by the SoHo™ implanted telemetry platform from Harvard Bioscience (HBio). This cutting-edge technology enables us to capture high-quality, continuous biopotential data of up to two channels, and includes precise activity and temperature recording in freely moving animals—opening new avenues for translational neuroscience and sleep research.
Figure 1: Workflow for sleep staging using SoHo™ implanted telemetry.
In a recent evaluation study, we deployed SoHo™ telemetry in the widely used 5xFAD mouse model of Alzheimer’s disease (AD) to investigate alterations in sleep architecture. By combining long-term electroencephalography (EEG) and electromyography (EMG) recordings with our automated, machine learning-based sleep staging pipeline, we achieved highly reliable classification of sleep states without the need for time-intensive manual scoring.
Our automated approach not only increases throughput and reproducibility but also enables deeper insights into sleep microstructure. In addition to standard sleep stage quantification, we performed detailed bout analysis, allowing us to characterize fragmentation patterns and state transitions with high temporal resolution. Our findings revealed phenotypic differences in the AD model, including increased wakefulness (Figure 2A, B), while REM and NREM sleep ratios remained unchanged (Figure 2D, E).
Figure 2: Automated sleep staging reveals increased wakefulness in 5xFAD mice. Animals were implanted with the recording device and after a recovery phase were monitored in the home cage for 7 days. Automated sleep staging algorithm classified sleep phases and data is shown averaged to 24h. Graphs show total time spent awake (A) and asleep (B) as well as percentage of sleep stages to total sleep time (TST) (C-E). Mean ± SEM; Wild type (WT) n = 4; 5xFAD n = 3; t-test; *p <0.05.
However, by doing in-depth bout analysis, where consecutive epochs are put in relation, we found alterations in REM sleep dynamics in the AD model, with reduced number of REM sleep events (Figure 3A) but significantly longer individual episodes (Figure 3D). Together, these features closely resemble the clinical manifestations observed in patients with Alzheimer’s disease.
Figure 3: Bout analysis elucidates REM sleep alterations in 5xFAD mice. Following sleep stage classification, consecutive epochs (bouts) were analyzed. Data is shown averaged over 24h. Graphs depict the number of bouts (A-C) and the average bout duration in minutes (D-F) per animal. Mean ± SEM; Wild type (WT) n = 4; 5xFAD n = 3; t-test; *p <0.05.
A key strength of our platform lies in its scalability and analytical depth. The integration of implanted telemetry with machine learning-driven data processing creates a streamlined, end-to-end workflow—from acquisition to interpretation—supporting consistent, unbiased analysis across large datasets. This makes the system particularly powerful for screening studies, longitudinal designs, and pharmacological interventions targeting sleep and neurological function.
While the primary focus of this study was on sleep architecture, the SoHo™ platform also enables comprehensive circadian rhythm analysis through continuous activity and core body temperature monitoring. Using established methods such as COSINOR and Non-Parametric Circadian Rhythm Analysis (NPCRA), we can quantify rhythmicity parameters including amplitude, phase shifts, and stability. Although not a central focus of this dataset, these capabilities highlight the platform’s versatility for future chronobiology investigations.
With these expanded capabilities, we are well-positioned to support a wide range of research applications—from neurodegenerative disease models and sleep disorders to stroke and circadian biology. We look forward to collaborating with partners to translate these insights into meaningful scientific and therapeutic advancements.
Additionally, please view the below video focusing on sleep disturbances in the 5xFAD mouse model, the often overlooked pathology reflects clinical observations and highlights the translational advantages of using implanted telemetry like the powerful SoHo™.
