My circle detection opencv program which calculates flow rate is getting stuck on an IndexError: "index 1 is out of bounds for axis 1 with size 1"

Question

My program is supposed to use circle detection in order to detect droplets falling through the drip chamber of an IV setup. It's then supposed to calculate the flow rate per minute based on the number of circles it detects. I got this code from a youtube video and mixed it with chatGPT. There is also an ROI in which the code focuses on in the video for better results. This has been bothering me for a while and I looked for fixed online but they are not working. the error is at the

 print(circles)
        # Adjust the circle coordinates based on the ROI
        chosen[:, 0] += ROI_X
        chosen[:, 1] += ROI_Y

section. Here is the full code:

import numpy as np
import cv2 as cv
import time

# Constants
PIXEL_TO_VOLUME_FACTOR = 0.0001  # Conversion factor from pixel count to volume (ml)
TIME_INTERVAL = 1.0  # Time interval between consecutive frames (seconds)

# Define the ROI coordinates and dimensions
ROI_X = 100  # x-coordinate of the top-left corner of the ROI
ROI_Y = 100  # y-coordinate of the top-left corner of the ROI
ROI_WIDTH = 20  # width of the ROI
ROI_HEIGHT = 20  # height of the ROI

capture = cv.VideoCapture(r'C:\Users\Jonah Anonuevo\Documents\SY 2022-2023\4th Quarter\Research\testvid.mov')
prevCircle = None
prevTime = time.time()
flow_rate = 0.0

while True:
    ret, frame = capture.read()

    # Crop the frame to the ROI
    roi_frame = frame[ROI_Y:ROI_Y + ROI_HEIGHT, ROI_X:ROI_X + ROI_WIDTH]

    grayFrame = cv.cvtColor(roi_frame, cv.COLOR_BGR2GRAY)
    blurFrame = cv.GaussianBlur(grayFrame, (17, 17), 2)

    circles = cv.HoughCircles(blurFrame, cv.HOUGH_GRADIENT, 1.2, 100, 100, 120, 5, 50, 100)

    if circles is not None and len(circles) > 0:  # Check if circles is not None and not empty
        circles = np.uint16(np.around(circles))
        chosen = circles[:2]  # Get the first set of circle coordinates

        print(circles)
        # Adjust the circle coordinates based on the ROI
        chosen[:, 0] += ROI_X
        chosen[:, 1] += ROI_Y

        # for c in chosen:
        cv.circle(frame, (c[0], c[1]), 1, (0, 100, 100), 3)
        cv.circle(frame, (c[0], c[1]), c[2], (255, 0, 255), 3)

        # Calculate flow rate
        current_time = time.time()
        elapsed_time = current_time - prevTime

        if prevCircle is not None:
            flow_rate = np.linalg.norm([chosen[0, 0] - prevCircle[0], chosen[0, 1] - prevCircle[1]]) * PIXEL_TO_VOLUME_FACTOR / elapsed_time
            prevTime = current_time

        #Display the flow rate
        print("Flow Rate: {:.2f} ml/s".format(flow_rate))

        prevCircle = chosen[0]  # Store only the first circle for comparison

    cv.imshow('Circles', frame)

    if cv.waitKey(1) == ord('q'):  # ends the loop when "q" is pressed
        break

capture.release()
cv.destroyAllWindows() 

After running, I get this:

PS C:\Users\Jonah Anonuevo\Documents\SY 2022-2023\4th Quarter\Research\Test Codes>  c:; cd 'c:\Users\Jonah Anonuevo\Documents\SY 2022-2023\4th Quarter\Research\Test Codes'; & 'C:\Program Files\Python311\python.exe' 'c:\Users\Jonah Anonuevo\.vscode\extensions\ms-python.python-2023.12.0\pythonFiles\lib\python\debugpy\adapter/../..\debugpy\launcher' '51286' '--' 'C:\Users\Jonah Anonuevo\Documents\SY 2022-2023\4th Quarter\Research\Test Codes\chatgpt circle detection.py'
[[100]
 [  0]
 [  0]
 [  0]]

I tried fixes online but none of them seem to work. Also tried asking my computer science-inclined friends for help but to no avail.