Rationale

Colonic high-resolution manometry (cHRM) is an emerging clinical tool for defining colonic function in health and disease. Current analysis methods are conducted manually, thus being inefficient and open to interpretation bias.

Objective

The main objective of the study was to build an automated system to identify propagating contractions and compare the performance to manual marking analysis.

Methods

cHRM recordings were performed on 5 healthy subjects, 3 subjects with diarrhea-predominant irritable bowel syndrome, and 3 subjects with slow transit constipation. Two experts manually identified propagating contractions, from five randomly selected 10-min segments from each of the 11 subjects (72 channels per dataset, total duration 550 min). An automated signal processing and detection platform was developed to compare its effectiveness to manually identified propagating contractions. In the algorithm, individual pressure events over a threshold were identified and were then grouped into a propagating contraction. The detection platform allowed user-selectable thresholds, and a range of pressure thresholds was evaluated (2 to 20 mmHg).

Key results

The automated system was found to be reliable and accurate for analyzing cHRM with a threshold of 15 mmHg, resulting in a positive predictive value of 75%. For 5-h cHRM recordings, the automated method takes 22 ± 2 s for analysis, while manual identification would take many hours.

Conclusions

An automated framework was developed to filter, detect, quantify, and visualize propagating contractions in cHRM recordings in an efficient manner that is reliable and consistent.

中文翻译：

---

ManoMap：用于表征由高分辨率测压记录的结肠传播收缩的自动化系统

基本原理

结肠高分辨率测压 (cHRM) 是一种新兴的临床工具，用于定义健康和疾病中的结肠功能。当前的分析方法是手动进行的，因此效率低下并且容易产生解释偏差。

客观的

该研究的主要目标是建立一个自动化系统来识别传播收缩并将性能与手动标记分析进行比较。

方法

对 5 名健康受试者、3 名患有腹泻型肠易激综合征的受试者和 3 名患有慢传输型便秘的受试者进行了 cHRM 记录。两名专家从 11 名受试者（每个数据集 72 个通道，总持续时间 550 分钟）中的每一个随机选择的五个 10 分钟片段中手动识别传播收缩。开发了一个自动信号处理和检测平台，以将其有效性与手动识别的传播收缩进行比较。在该算法中，识别出超过阈值的单个压力事件，然后将其分组为传播收缩。检测平台允许用户选择阈值，并评估了一系列压力阈值（2 到 20 mmHg）。

主要结果

发现该自动化系统在分析 cHRM 时可靠且准确，阈值为 15 mmHg，阳性预测值为 75%。对于 5 小时的 cHRM 记录，自动化方法需要 22 ± 2 s 进行分析，而手动识别需要很多小时。

结论

开发了一个自动化框架，以可靠且一致的有效方式过滤、检测、量化和可视化 cHRM 记录中的传播收缩。