chg: update project page

index.md

@@ -2,22 +2,73 @@
 layout: posts
 title:  "Continuous Patient Monitoring with AI: Real-Time Analysis of Video in Hospital Care Settings"
 ---
+> **Authors**: Paolo Gutierrez Gabriel (corresponding), Peter Rehani, Tyler Troy, Tiffany Wyatt, Michael Choma and Narinder Singh
+>
+> _The manuscript is submitted for review in 'Frontiers in Imaging - Imaging Applications'._
 
-> NOTE [2024-12-17]: this repository and project page are under construction
+## Welcome! 
+You are likely here because you read the manuscript, "Continuous Patient Monitoring with AI: Real-Time Analysis of Video in Hospital Care Settings", and are looking for the public data and associated code that was mentioned. 
 
-**Authors**: Paolo Gutierrez Gabriel, Peter Rehani, Tyler Troy, Tiffany Wyatt, Michael Choma and Narinder Singh
+![Figure 1 of pre-print](assets/images/overview.png "LookDeep Health System")
+_Figure 1 of pre-print article_
 
-This is the corresponding code and data repository for the manuscript titled "Continuous Patient Monitoring with AI: Real-Time Analysis of Video in Hospital Care Settings".
+Thank you for your interest in analyzing these anonymized computer vision predictions, aggregated at the patient-day-hour level. 
+We invite you to explore the trends reported in the manuscript (and more), and provide the materials below as support.
 
-The manuscript is submitted for review in *Frontiers in Imaging
-Imaging Applications*.
-* [arXiv pre-print](https://arxiv.org/abs/2412.13152) 
+> **_NOTE_** [2024-12-17]: this repository and project page are under construction. 
+thank you for your patience. 
+
+### Dataset
+
+This project page is about a new, fully anonymized dataset (unnamed).
+This data consists of computer vision predictions recorded from hundreds of patients over a 6-month period, aggregated at the patient-day-hour level. 
+
+![Example trend](assets/images/trends_plot.png "Percent Time Alone")
+
+_Example trend plot, rendered in [Vercel App](https://lookdeep.vercel.app/)_
 
-The corresponding repo is maintained on the `main` branch, and can be found at [[link](https://github.com/lookdeep/ai-norms-2024)]. 
+These trends provide hour-over-hour insight into various patient behvaiors and room conditions, including but not limited to:
+```
+**Metadata:**
+- model_name
+- day_loc
+- hour_loc
+- monitor_id
+- group_ADC
+- gender
+- group_age
+- ts
+
+**Data:**
+- num_samples
+- percent_alone
+- count_fence_crossed
+- percent_fence_crossed
+- percent_bed_active
+- percent_scene_active
+- count_alone_active
+- percent_alone_active
+- count_alone_fence_crossed
+- percent_alone_fence_crossed
+- percent_staff
+- percent_supervized
+- percent_supervized_staff
+- percent_supervized_other
+- percent_supervized_staff_fence_crossed
+
+```
+
+*Details*:
+- direct link to data: lookdeep/ai-norms-2024/data_by_patient_hour_scrubbed.csv
+- computer vision predictions come from Models v4 and v5 -> see manuscript for evaluation
+
+## Resources
+* [arXiv pre-print](https://arxiv.org/abs/2412.13152) 
+* [code repository](https://github.com/lookdeep/ai-norms-2024) - use `main` branch
+* [LookDeep Health website](https://lookdeep.health/technology/)
 
 ---
 \
-**To-do**:
- * [] Update project page
- * [] Add public dataset
- * [] Add demo code 
+**Change Log**:
+ * 2024-12-17 : initialize project page and link to pre-print
+ * 2024-12-18 : update project page with dataset