# uses Dijkstra's algorithm to find minimal path

infinity = 999999

matrix = []

width = 80
height = 80

f = open('081.txt', 'r')
for line in f:
	line.rstrip('\n')
	line_str = line.split(',')
	line_int = [ int(x) for x in line_str ]
	matrix.append(line_int)
f.close()


# initialize
distance = [infinity] * (width*height)
pre = [None] * (width*height)
distance[0] = 0
Q = range(width*height)

def mindistance(Q):
	mind = infinity
	mini = -1
	for i in Q:
		if distance[i] < mind:
			mind = distance[i]
			mini = i
	return mini

def neighbors(u):
	n = []
	row = u / width
	col = u % width
	if col < width-1:
		n.append(row*width + col + 1) # right neighbor
	if row < height-1:
		n.append((row+1)*width + col) # lower neighbor
	return n

def update(u, v):
	alt = distance[u] + matrix[v/width][v%width]
	if alt < distance[v]:
		distance[v] = alt
		pre[v] = u

# main algorithm
while len(Q) > 0:
	u = mindistance(Q)
	Q.remove(u)
	for v in neighbors(u):
		if v in Q:
			update(u, v)

# search min distance
u = width*height-1  # target node
path = [ u ]
while pre[u] != None:
	u = pre[u]
	path = [u] + path

minsum = 0
for n in path:
	minsum += matrix[n/width][n%width]

print minsum